Issues and Methods in the Detection of Alcohol and Other Drugs

TRANSPORTATION RESEARCH

E-CIRCULARNumber E-C020 September 2000

Issues and Methods in theDetection of Alcohol and

Other Drugs

TRANSPORTATION Number E-C020, September 2000RESEARCH ISSN 0097-8515E-CIRCULAR

Issues and Methods in the Detection of Alcohol and Other Drugs

COMMITTEE ON ALCOHOL, OTHER DRUGS, AND TRANSPORTATION

Allan F. Williams, Chair

Marcelline Burns Hans Laurell Stephen M. SimonCarol P. Council Susan E. Martin Herbert M. Simpson

James Fell J. J. McLean Carl A. SoderstromSusan A. Ferguson Judy Z. Meade Barry M. Sweedler

Robert D. Foss Herbert Moskowitz Evelyn VingilisJames H. Hedlund Wolf-Ruediger Nickel Robert B. VoasSusan B. Herbel Raymond C. Peck J. Michael WalshRalph Hingson David F. Preusser Elisabeth N. Wells-ParkerBrian A. Jonah Jean T. Shope

Richard F. Pain, TRB Staff RepresentativeRobert Hilterbrand, TRB Senior Program Assistant

Subscriber categories Transportation Research BoardIIA highway and facility design National Research Council4B safety and human performance 2101 Constitution Avenue, NWnationalacademies.org/trb Washington, DC 20418

The Transportation Research Board is a unit of the National Research Council, which is the principal operating agency of theNational Academy of Sciences and the National Academy of Engineering. The research council provides independent advice onscientific and technical matters under a congressional charter granted to the National Academy of Sciences, a private, nonprofitinstitution dedicated to the advancement of science and technology and to their use for the general welfare.

The Transportation Research Board is distributing this Circular to make the information contained herein available for use byindividual practitioners in state and local transportation agencies, researchers in academic institutions, and other members of thetransportation research community. The information in this Circular was taken directly from the submissions of the authors. Thisdocument is not a report of the National Research Council or of the National Academy of Sciences.

http://gulliver.trb.org/publications/circulars/NASpage.pdf

FOREWORD

This TRB circular includes the papers and discussion from a midyear 1998 meeting andworkshop organized by the Committee on Alcohol, Other Drugs, and Transportation(A3B10). The conference chair and organizer was Allan F. Williams. James H. Hedlundprepared this circular. In the workshop, the presentation of one or more discussion papers,or comments, followed each paper presentation. Then questions and answers and a generaldiscussion period provided all participants the opportunity to explore the topic further andexpress opinions on research or program needs, ideas, or opportunities.

CONTENTS

Workshop Introduction:ISSUES AND METHODS IN THE DETECTION OF ALCOHOL AND OTHER DRUGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1Allan F. WilliamsInsurance Institute for Highway SafetyRobert B. VoasPacific Institute for Research and EvaluationSusan A. FergusonInsurance Institute for Highway Safety

Workshop Summary:WHAT’S NEEDED TO IMPROVE POLICE DETECTION OFALCOHOL AND OTHER DRUGS IN DRIVERS . . . . . . . . . . . . . . . . . . . . . . . . B-1James HedlundHighway Safety North

LEGAL AND CONSTITUTIONAL ISSUES RELATED TO DETECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-1Michele FieldsInsurance Institute for Highway Safety

INCREASING THE OPPORTUNITIES TO EXAMINE IMPAIRED DRIVERS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-1Jack W. StusterAnacapa Sciences, Inc.

Comments on “Increasing the Opportunities to ExamineImpaired Drivers” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-16Douglas J. BeirnessTraffic Injury Research Foundation

Comments on “Increasing the Opportunities to ExamineImpaired Drivers” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-19Barry M. SweedlerNational Transportation Safety Board

Comments on “Increasing the Opportunities to ExamineImpaired Drivers” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D-22James C. FellNational Highway Traffic Safety Administration

IDENTIFICATION OF ALCOHOL IMPAIRMENT ONINITIAL INTERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-1David F. PreusserPreusser Research Group, Inc.

Comments on “Identification of Alcohol Impairment onInitial Interview” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-8Bill TowerMaryland State Police Department

Comments on “Identification of Alcohol Impairment onInitial Interview” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-11Susan A. FergusonInsurance Institute for Highway Safety

IDENTIFICATION OF IMPAIRMENT OUTSIDE THE VEHICLE:FIELD SOBRIETY TESTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F-1Marcelline M. BurnsSouthern California Research Institute

Comments on “Identification of Impairment Outside the Vehicle” . . . . F-13Robert B. VoasPacific Institute for Research and Evaluation

Comments on “Identification of Impairment Outside the Vehicle” . . . . F-20James HedlundHighway Safety North

EVIDENCE OF ALCOHOL AND DRUG IMPAIRMENT OBTAINED AFTER ARREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G-1Steve SimonUniversity of Minnesota

Comments on “Evidence of Alcohol and Drug Impairment Obtained After Arrest” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G-15Herbert A. MoskowitzSouthern California Research Institute

Comments on “Evidence of Alcohol and Drug Impairment Obtained After Arrest” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G-19Michele FieldsInsurance Institute for Highway Safety

ISSUES IN THE DETECTION OF DRUGS OTHER THAN ALCOHOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H-1J. Michael WalshThe Walsh Group, P.A.

Comments on “Issues in the Detection of Drugs Other Than Alcohol” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H-11Han de GierUniversity of Utrecht

Comments on “Issues in the Detection of Drugs Other Than Alcohol” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H-14Richard P. ComptonNational Highway Traffic Safety Administration

PROSECUTING AND ADJUDICATING DWI DETECTION EVIDENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I-1Joel A. WatneState of Minnesota

Comments on “Prosecuting and Adjudicating DWIDetection Evidence” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I-25Kathryn StewartPacific Institute for Research and Evaluation

Comments on “Prosecuting and Adjudicating DWIDetection Evidence” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I-27William G. SlevaLawrence Superior Court II

WORKSHOP INTRODUCTION

Issues and Methods in the Detection ofAlcohol and Other Drugs

ALLAN F. WILLIAMS

Insurance Institute for Highway SafetyROBERT B. VOAS

Pacific Institute for Research and EvaluationSUSAN A. FERGUSON

Insurance Institute for Highway Safety

INTRODUCTION

The central question of this workshop is: How can it reliably be established that a personhas consumed alcohol and/or other drugs in amounts that are illegal and/or may causeharm to that person or others? Many people think of this as a relatively easy task,conjuring up images of the “falling down drunk,” although even this obvious display maybe something else, for example, diabetic shock. The reality is that even when people havehigh blood alcohol concentrations (BAC) known to cause significant impairment, it canbe difficult for experts and laypersons alike to detect alcohol, especially among seasonedusers and among people who wish to remain undetected. Detecting drugs other thanalcohol presents its own set of problems, and in many cases the behavioral cues are lessobvious than when alcohol has been consumed. The workshop’s focus is on policedetection of alcohol and other drugs among drivers of private motor vehicles.

DETECTION PROBLEMS

There are many examples that illustrate serious detection problems involving alcoholconsumption. In a 1957 study in which medical assessments of impaired driving abilitywere compared with results of actual driving tests, it was concluded that “a medicalexamination alone is not a reliable means of detecting alcoholic impairment of drivingability” (Crime Detection Laboratory, 1957). In several investigations around the worldin which people at various BACs were diagnosed as drunk or not, only 62 percent ofpeople with BACs of 0.10–0.15 percent were thought to be drunk (American MedicalAssociation, 1968).

Tolerance to alcohol allows many heavy drinkers to escape detection, even at veryhigh BACs (Chesher and Greeley, 1992). Rosen and Lee (1976) found that althoughsocial drinkers exhibited behavioral signs of intoxication at BACs of 0.10 percent,alcoholics showed virtually none, even though both groups were equally impaired oncognitive performance measures involving recall of lists of numbers and words. Perper et al. (1986) reported that many alcoholics admitted to a detoxification unit had normalspeech, gait, and unimpaired ability to undress, even with BACs of 0.35 percent andgreater. At sobriety checkpoints conducted in North Carolina, 60 percent of drivers with

A-1

A-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

BACs of 0.10 percent or greater were not detained by police for further testing (Wells et al., 1997). In a study of the detection of alcohol under ideal laboratory conditions, 40 percent of drinking subjects with BACs greater than 0.08 percent were not identifiedby alcohol odor after they had eaten some food (Moskowitz et al., 1997).

There is scant literature on the detection of drugs other than alcohol amongdrivers or in other situations. On the road, drivers sometimes come under suspicion ofhaving ingested drugs when they act impaired but test negatively for alcohol. In thiscontext, police officers trained in drug recognition techniques are reasonably accurate indetermining if drugs are, in fact, involved (Preusser et al., 1992). However, it is notknown how many drugged drivers there are who do not come under suspicion in thismanner.

Most of the examples involving alcohol detection pertain to identifying peoplewith high BACs. However, depending on the target population, it is also necessary toidentify people with lower BACs—0.08 percent, 0.04 percent, or any alcohol—and this iscorrespondingly more difficult.

IMPORTANCE OF DETECTION IN VARIOUS SITUATIONS

Although the workshop focuses on police detection of alcohol and other drugs amongdrivers, there are many other instances in which we wish to detect alcohol and otherdrugs: commercial transportation in all forms; non-motor-vehicle pursuits such asrecreational boating; industry; medical settings in cases in which it may be necessary todistinguish alcohol intoxication from head injury for treatment purposes; school settings;treatment settings to make sure clients are following treatment regimens; and retailestablishments where there are alcohol servers.

Detection of alcohol and other drugs in these other realms differs from what isinvolved in dealing with private motor vehicle drivers. For example, in some casesmandatory testing is involved on a random basis or prior to performing the activity. Asthe result of the 1991 Omnibus Transportation Employee Testing Act, every driver of alarge bus or truck, or anyone who is transporting people commercially, is subject torandom alcohol and drug tests. Also, individuals attending drug and alcohol treatmentprograms, often as the result of a DWI conviction, are subject to testing at the discretionof the counselor. For alcohol servers in retail establishments, the issue is usually notwhether patrons have been drinking, but whether servers should continue to serve themdrinks.

The situation most analogous to the detection of alcohol and drugs on thehighway is the detection of impaired boat operators. In most states, the laws governingdrinking and operating boats are the same ones governing driving motor vehicles on theroad, and detection of boat operators with illegal BACs by necessity follows similarpaths to that for operators of motor vehicles. For example, marine police might stopboats that are being operated in a risky manner, such as powerboats being operated athigh speeds at night. They also might examine more closely operators who have beeninvolved in a mishap. However, until recently there has been little effort tosystematically deter this population from drinking. Research now is under way to try tounderstand the magnitude of the problem and to develop field sobriety tests more suitedto on-board administration.

Williams, Voas, and Ferguson A-3

DETECTION ON THE HIGHWAY

The detection process on the highway involves two stages: accurately detecting alcoholor other drugs and establishing the basis for criminal prosecution. That is, even though apolice officer may successfully detect illegal use of alcohol or other drugs, there aredetection-related issues that may hinder the application of sanctions. The workshopaddresses both of these issues and emphasizes that detection and follow-up processesmust be done in the context of maintaining the legal and constitutional rights ofindividuals. Notably, the U.S. Constitution prohibits some detection techniques usedadvantageously in other countries.

EVOLUTION OF DETECTION ISSUES

The workshop addresses detection in regard to current law and practice, but it is ofinterest that there have been changes in laws against alcohol-impaired driving over theyears that have changed the way in which the process of detection and building a caseproceed. During the first half of this century, evaluating the degree of impairment of adriver who had been drinking focused principally on the description of behavior. The firststate laws prohibited driving while intoxicated or under the influence of alcohol. Inpractical terms, this meant that only obviously impaired drivers were likely to be arrested.In fact, as research subsequently determined, many alcohol-impaired drivers do notappear drunk, and at that time it was difficult to obtain a conviction because no objectivestandard existed to prove intoxication. Following World War II, the nation began theprocess of integrating chemical testing with DUI enforcement. At first, this was in theform of a blood or urine test based on the work of Widmark early in the century. Initially,this was done through what are known as presumptive laws, which establish apresumption of impairment at or above a specified BAC (defendants could try to rebut thepresumption). The most significant advance in chemical alcohol test technology was thedevelopment of a practical breath test by Borkenstein, opening the way to its widespreaduse throughout the United States in the 1960s. Reliable chemical testing permitted theadoption of per se laws, first adopted by Norway in 1936, that define the offense asdriving with a BAC above a proscribed limit. In the United States, per se laws were notintroduced until the 1970s but now are in place in 49 states and the District of Columbia.Defendants charged with per se offenses can no longer try to prove they were notimpaired, although they can challenge the validity of the BAC test. Even with per selaws, however, behavior plays a vital role in the arrest and conviction of impaired driversin the United States in two ways. In the field, behavior provides justification for the DUIinvestigation and arrest, which in turn provides the officer with the authority to requirethe breath test.

WORKSHOP FOCUS

The workshop traces the detection process from the moment the person enters a vehiclethrough to the sanctioning process. It should be noted, however, that the detection processcan start earlier than this. For example, alcohol servers or passengers may detectimpairing amounts of alcohol and deter potential drivers from driving. The workshop

A-4 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

does not attempt to cover all aspects and procedures involved in the arrest process butemphasizes those involving detection and the proof that what has been detected is illegal.

Detection is a relevant topic because it is a key to deterring impaired driving, butit is not easy to accomplish. If people do not think there is much of a chance they will beencountered by the police, or if encountered still are not likely to be detected, or ifdetected are not likely to be successfully sanctioned, they will be little discouraged fromdriving while impaired. The goals of the workshop are to examine the detection process,see how it might be improved, and see what the research needs are.

PAPERS PRESENTED

The first paper, by James Hedlund, was written after the workshop took place andbasically summarizes key points, issues, and ideas from the workshop.

Papers presented at the workshop address, in sequence, the various stages inencountering and sanctioning persons suspected of alcohol-impaired driving. The paperby Michele Fields sets the stage, laying out the legal/constitutional constraints in theprocess, the rules of the game that govern how detection of illegal impairment proceedsin the United States while protecting the rights of individuals.

In the second paper, Jack Stuster discusses ways to increase opportunities forpolice to examine impaired drivers. The percentage of impaired drivers on the road whoare detected is very low; one important reason for this is that few impaired motorists areever encountered by the police in the first place. Stuster discusses various ways toincrease the intersection of police and impaired drivers and to recognize vehicles thatcontain impaired drivers.

Then come three papers that trace the buildup of evidence that may lead to arrestonce a vehicle is stopped. David Preusser discusses detection techniques and issues thatcome into play in the at-the-window encounter with the driver. Marcelline Burnsdiscusses identification of impairment once the driver has been asked to exit the vehicle.Steve Simon discusses issues and procedures involving detection once an arrest has beenmade, focusing on evidential testing issues.

The workshop addresses both alcohol and other drugs, but much of the researchdiscussed in the papers is based on alcohol alone. Michael Walsh specifically addressesissues and techniques in the identification of drug impairment at the roadside and in thepolice station, including behavioral cues, drug recognition training, screening andchemical tests, and issues in determining impairment.

In the final paper, Joel Watne discusses prosecution and adjudication issuesrelated to detection, basically addressing the questions as to how detection evidence getschallenged, how people get off, and what changes might improve the detection process.

REFERENCES

American Medical Association. Alcohol and the Impaired Driver. Chicago, Ill., 1968.

Chesher, G., and J. Greeley. Tolerance to the Effects of Alcohol. Alcohol, Drugs, and Driving, Vol. 8, No. 2,1992, pp. 93–106.

Crime Detection Laboratory (Royal Canadian Mounted Police). Report on Impaired Driving Tests(B. B. Coldell, ed.). Ottawa, Canada, 1957.

Williams, Voas, and Ferguson A-5

Moskowitz, H. A., M. M. Burns, and S. A. Ferguson. Detecting Breath Odor from Alcohol Consumption.Proceedings of the 14th International Conference on Alcohol, Drugs, and Traffic Safety, Vol. 2, 1997,pp. 527–531. CERMT, Annecy, France.

Perper, J. A., A. Twerski, and J. W. Wienand. Tolerance at High Blood Alcohol Concentrations: A Studyof 110 Cases and Review of the Literature. Journal of Forensic Sciences, Vol. 31, No. 1, 1986, pp. 212–221.

Preusser, D. F., R. G. Ulmer, and C. W. Preusser. Evaluation of the Impact of the Drug Evaluation andClassification Program on Enforcement and Adjudication. U.S. Department of Transportation,Washington, D.C., 1992.

Rosen, L. J., and C. L. Lee. Acute and Chronic Effects of Alcohol Use on Organizational Processes in Memory.Journal of Abnormal Psychology, Vol. 85, 1976, pp. 309–317.

Wells, J. K., M. A. Greene, R. D. Foss, S. A. Ferguson, and A. F. Williams. Drinking Drivers Missed atSobriety Checkpoints. Journal of Studies on Alcohol, Vol. 58, No. 5, 1997, pp. 513–517.

WORKSHOP SUMMARY

What’s Needed to Improve PoliceDetection of Alcohol and Other

Drugs in DriversJAMES HEDLUND

Highway Safety North

INTRODUCTION

This workshop examined the apparently narrow and well-defined topic of police detectionof alcohol and other drugs among drivers of private motor vehicles. The workshop appearedto be limited to impaired driving enforcement: How police identify drivers who may beimpaired, and how they determine whether drivers are under the influence of alcohol andother drugs. But workshop participants quickly came to appreciate that these enforcementissues were intertwined with the entire system of impaired driving laws, enforcement, andsanctions. Consequently, the workshop papers, discussions, and ideas had a broad range.

This paper summarizes some key points made at the workshop. It concentrates onthe workshop’s goals. How can police detection be improved? What changes in legislationand programs make any difference?

This paper attempts to capture the sense of workshop participants as expressed intheir papers and discussions. Statements in this paper are the author’s interpretation ofgeneral (though not necessarily unanimous) thinking by workshop participants. Pointsmade at the workshop that may not have had broad support are described as views of“some participants.” Because only authors of papers reviewed this summary, its statementsmay not represent a majority view. Of course, any workshop participant (including theauthor) may disagree with any of the statements in this paper.

Throughout this paper, the term “impaired drivers” refers to drivers impaired byalcohol or other drugs. “Alcohol-positive” refers to drivers with any measurable positiveblood alcohol content. Similarly, “drug-positive” refers to drivers with a measurableamount of any drug that can impair driving performance. “Police” refers to any lawenforcement officer with the authority to enforce impaired driving laws.

WHY DETECT IMPAIRED DRIVERS?

The system of laws, enforcement, prosecution, and sanctions directed at impaired driversexists both to catch impaired drivers and, more important, to deter impaired driving: toconvince people not to drive while impaired because they will suffer consequences if theydo. Deterrence theory says that to be effective, these consequences must be swift, certain,and severe (and research strongly suggests that swiftness and certainty are more importantthan severity). The impaired driving system already provides for sanctions that are tovarying degrees swift and severe for impaired drivers. But they are far from certain.

B-1

B-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

Recent estimates suggest that fewer than 1 impaired driving trip in 500 will be detected bypolice. This defines the workshop’s goal: to improve police detection, which will increasethe likelihood that impaired drivers will be caught and punished and in turn deter othersfrom driving while impaired by alcohol or other drugs.

Participant Discussion of research and experience concluded that the priority andresources allocated to law enforcement by the community and its leaders are keyingredients in detecting impaired drivers. Police need to be on the street and know thatdetecting impaired drivers is an important part of their duties. Lacking this, the programand ideas that follow will have little effect. One important way to influence communitypriorities is through continued public education on all aspects of impaired driving.

THE LEGAL SYSTEM: IMPAIRED DRIVING LAWS AND PRACTICES

Police detection of impaired drivers operates within the system of laws and practices thatdefine, enforce, prosecute, and adjudicate traffic offenses. The laws define the variousimpaired driving offenses in the first place. The practices establish what police can andcannot do in stopping, investigating, arresting, and obtaining evidence from suspectedimpaired drivers. In the United States, each of the 50 states and the District of Columbiahas its own set of laws and practices. The workshop’s considerations apply to thesejurisdictions in the United States. Participants from other countries added usefulcomments on the U.S. situation from an international perspective.

State impaired driving laws come in two forms: behavioral and chemical. All stateshave “behavioral” laws, under which it is illegal to drive while impaired by alcohol orother drugs. Conviction under these laws requires police to obtain evidence that thedriver’s performance was impaired and to link this impairment to alcohol or other drugs.Evidence that the driver was alcohol- or drug-positive is not by itself sufficient forconviction. All states also have “chemical,” or per se, laws for alcohol, under which it isillegal to drive with a blood alcohol concentration (BAC) exceeding a specified level (twostates have per se laws only for drivers younger than age 21). Some states also have per selaws for certain illicit drugs. However, in practice, BAC or drug evidence alone is notsufficient to win conviction under a per se law (much to the surprise of foreign workshopparticipants, where BAC evidence by itself will suffice). Workshop participants familiarwith driving while impaired (DWI) trials in several states confirmed that BAC evidence isimportant, and, in fact, convictions are difficult without a valid BAC reading above thestate’s per se level, but BAC evidence must be accompanied by behavioral evidence ofimpairment. Some participants mused that detection would be simplified considerably ifthis were not so, but Participant Discussion clearly pointed out that this situation isunlikely to change in the foreseeable future. Consequently, police must acquire bothbehavioral and chemical evidence of impairment.

Within this basic framework, state impaired driving laws regarding both alcoholand drugs have developed an almost Byzantine structure of offenses, penalties, andexceptions. As the laws and procedures become more complex, successful convictionbecomes more difficult. Workshop participants noted that the basic model for these laws,the impaired driving section of the Uniform Vehicle Code, has not been reviewed formany years. Participant Discussion pointed out the need for a review, with the goal ofupdating and simplifying the code.

Hedlund B-3

Police detection practices also must operate within the constraints imposed by theUnited States and individual state constitutions that protect citizens from unreasonablesearches and seizures and provide for due process and other rights. The key issues arethat police require “articulable suspicion” of some traffic offense in order to make atraffic stop: Except at sobriety checkpoints, they cannot arbitrarily stop a car for noreason (as police in some other countries can). Once a car is stopped, police again cannotrequest a driver to step out of the vehicle or provide a breath sample unless they againhave some articulable suspicion of alcohol or other drugs. (A few states allow police torequest a breath sample from a driver who has committed a moving traffic violation orwho has been involved in a crash.) And finally, to make an impaired driving arrest, policemust have a higher level of evidence: “probable cause” of the offense.

These constraints have given rise to what is in many states an elaborate system ofprocedures for impaired driving enforcement and arrest. On the one hand, following clearlyestablished procedures can prevent challenges on the grounds of arbitrary behavior from thepolice. But on the other hand, as procedures become more elaborate, more police time isrequired for every impaired driving arrest and the chances are greater that police willinadvertently fail to follow the procedures precisely. Any deviation can be challenged incourt and may cause the case to be lost. Workshop participants noted the potential benefits ineffectiveness and efficiency from review and simplification of these procedures in all states.

Administrative sanctions for impaired driving can bypass some of these criminaljustice system issues. Most states have adopted administrative license revocation orsuspension (ALR) for drivers who violate the state’s per se law. ALR laws have beenshown to reduce impaired driving. Administrative license plate suspension also is used.As participants noted, these and other administrative measures are proven techniquesavailable now for states to conduct.

The needs are as follows:

• An updated Uniform Vehicle Code section on impaired driving;• More uniform state impaired driving laws;• Simplified state impaired driving arrest and evidence procedures; and• Broader use of administrative sanctions.

ON THE ROAD: OBSERVING AND IDENTIFYING DRIVERS WHO MAY BE IMPAIRED

Impaired driving detection begins with police observing a driver and observing a reasonto stop him or her. The reason can be a traffic infraction, an action that is not in itself aninfraction but is associated with impaired driving (such as driving unreasonably slow), ora crash. Police detect impaired drivers in three basic ways, each of which presents its ownchallenges and opportunities.

Traffic Patrol

Most police traffic time is spent in traffic patrol activities, and most impaired drivingarrests result from patrol. Patrol includes routine traffic patrol, special impaired drivingenforcement such as “blanket” or “saturation” patrols, and other traffic activities such as


special event traffic control. In some of these activities, police are looking specifically forimpaired drivers; in others, they are not.

Clearly, the more time police are on the road, the greater are their chances ofobserving impaired drivers. Thus, the first method to improve impaired driving detection isto increase police patrol time. One obvious way is to increase the number of patrol officers.This, of course, either requires additional funds or else it diverts police resources from otherneeds. A second way is to reduce patrol officer “down time” so that they can spend morehours on the road. A major contributor to this is impaired driving arrests themselves,which may take as long as four hours of an officer’s time. Workshop discussion stronglysupported measures to reduce impaired driver processing time through such means assimplified paperwork requirements and automated record systems. Why, for example,must police in one state fill out eight separate forms, each containing some of the sameinformation, when a computerized system would allow information to be entered only once(and could check for some obvious errors)?

The second method to improve impaired driving detection is to help police usetheir patrol time more efficiently, by providing information on where patrol officers shouldlook and what should they look for. Some patrol activity is targeted to locations and hoursfavored by impaired drivers (e.g., near bars on weekend nights). Some jurisdictionsregularly patrol locations where convicted impaired drivers live (identified on “HOTsheets”). Patrol officers also use a number of established cues to identify drivers who maybe impaired. Participants identified two areas where improvements could be made. First,police generally are unfamiliar with cues to identify young drivers who may be impaired,because young impaired drivers are found in different places at different times and exhibitdifferent driving behaviors than older impaired drivers. Second, if police could accessvehicle license plate and driver records from their patrol cars, they could determinequickly if a vehicle is licensed to a convicted impaired driver. When a person’s credit cardinformation is available within seconds at virtually every gas station and convenience storein the country, why cannot driver’s license and driver record information be equallyaccessible to police?


• Increased patrol time by reducing and simplifying impaired driver processingrequirements and paperwork;

• A handbook of impaired driving patrol strategies and tactics;• Patrol strategies and cues for detecting youthful impaired drivers; and• Improved police access to vehicle and driver records from the patrol car.

Checkpoints

At a sobriety checkpoint, police stop every vehicle (or every second or third vehicle) at apredetermined location to check whether the driver is impaired. Checkpoints can be aneffective way of deterring impaired driving because they require all drivers to stop:Experienced and alcohol-tolerant impaired drivers cannot avoid being stopped by drivingcarefully. Extensive and well-publicized checkpoint programs have been shown to reduceimpaired driving crashes.

Checkpoints present three major issues: their legal authority in some jurisdictions,the resources required to conduct them, and their effectiveness in detecting impaired

Hedlund B-5

drivers. Checkpoints are permitted in 39 states and the District of Columbia. The U.S.Supreme Court has found that checkpoints are constitutional, but some state courts haveheld that checkpoints violate state constitutions. In other states, checkpoints would bepermitted if authorized by the state legislature, but the legislature has not yet acted. Somecheckpoints are conducted with many police officers and consequently impose a heavytoll on police resources. Recent research has shown that smaller checkpoints can be justas effective as larger ones. Workshop participants encouraged additional research andguidelines on the most cost-effective ways to operate checkpoints. Finally, research hasshown that approximately one-half of all legally impaired drivers stopped at checkpointsare not detected. In addition to the points discussed in the following section (At the CarWindow), research on cues to identify drivers entering checkpoints by using their vehiclemovements also could help police at checkpoints identify drivers who might be impaired.


• Wide use of checkpoints in states where checkpoints are constitutional but not yetpermitted;

• Research and guidelines on cost-effective checkpoint operation; and• Research on cues to use vehicle movements at checkpoints to identify drivers who

might be impaired.

Citizen Reports

Some jurisdictions have a process for citizens to report suspected impaired drivers byusing a special telephone or cell phone number. This method of reporting suspiciousbehavior could increase substantially as more and more drivers have cell phones in theirvehicles. But it will be useful only if police have the resources to monitor and follow upthese reports.


• Research on procedures for and the effectiveness of reporting suspected impaireddrivers by cell phone.

AT THE CAR WINDOW: OBTAINING FURTHER EVIDENCE

After a vehicle is stopped, whether in patrol operations or at a checkpoint, police mustdetermine quickly whether there is enough evidence of impairment to require the driver tostep out of the car for further investigations. Sometimes this is easy: There may be directphysical evidence of alcohol or drugs (smell, open containers, etc.), or the driver mayadmit to alcohol or drug use in response to the officer’s questions. Lacking these, theofficer must attempt to uncover any evidence in other ways.

Many experienced officers have developed strategies to check for impairment.One promising method is to examine the driver’s eyes for horizontal gaze nystagmus(HGN). HGN is highly correlated with impairment by alcohol and some drugs and is anintegral part of the Standardized Field Sobriety Tests (SFST), but it has not been usedwidely at the vehicle window. Workshop participants encouraged research to see if HGNshould be taught to patrol officers for use at the car window.


Technology can help. The passive alcohol sensor (PAS) can determine whetheralcohol is present in the occupant compartment. The PAS can be especially useful atcheckpoints, where police must make a quick decision based on almost no other evidencebeyond what they observe at the car window. Participant Discussion supported continueddevelopment and broader use of the PAS.

Young drivers again present special challenges. Police need cues and strategiesfor interviewing youth, especially for enforcing zero tolerance laws with a BAC limit of0.02 percent or less (now in effect in every state). A PAS could be particularly useful inthe absence of obvious behavioral cues.


• Research on the use of HGN at the vehicle window as an initial indication ofimpairment;

• Increased PAS use in patrol and checkpoint operations;• Strategies for police to enforce zero tolerance laws for youth; and• A PAS for zero tolerance law enforcement.

Crashes

Drivers in crashes already have given police a reason to question them and to check forpossible impairment. However, police investigating a crash have many pressingresponsibilities. Some participants suggested that research to develop interview protocolsfor police at the crash scene may help identify impaired drivers. Others suggested that allcrash-involved drivers be tested for alcohol or drug impairment (as is permitted in threestates).

Injured drivers present special problems. Because the first concern of police at acrash scene is to treat any injuries, these drivers usually are taken directly to hospitals oremergency rooms. Although the treatment facility often acquires a blood sample for usein treatment, the blood may not be tested for alcohol or drugs; if tested, results may notbe given to police. This issue raises broad questions of physician and hospital roles andobligations to their patients on the one hand and to society as a whole on the other hand.


• Police protocols for interviewing crash-involved drivers;• Breath testing crash-involved drivers (if permissible under state law); and• Research methods to acquire BAC results for injured drivers.

OUTSIDE THE VEHICLE: OBTAINING EVIDENCE TO MAKE AN ARREST

After the driver has stepped outside his vehicle, the officer has some evidence pointing toimpairment. The officer’s task now is to decide whether there is enough evidence for anarrest. If so, the officer also must document evidence that can be used in court. As at thecar window, both behavioral and technological methods can be used.

The SFST consists of three behavioral tests: walk and turn, one-leg stand, and HGN.These tests were developed to determine whether a driver’s BAC exceeded 0.10 percent.Recent research again validated the SFST and showed that it is equally valid at predicting

Hedlund B-7

a BAC of 0.08 percent. The SFST is used to some extent in all states and is taught routinelyto police officers in police academies. The SFST, especially HGN, has been challenged incourt and has been upheld in most, but not all, decisions.

The key issue with the SFST is to ensure the approximately 400,000 patrolofficers in the United States are proficient in its use. In addition, efforts must continue toeducate the courts on the SFST and to win legal challenges to its validity.

Technology again can help. A preliminary breath test (PBT) device measures thealcohol content in a driver’s breath. Approximately half the states now allow PBTs to beused to provide evidence for an arrest (but not to establish the driver’s BAC).

Workshop participants also mentioned several methods to acquire additionalinformation or help prepare for trial. Once again, if the officer has online access to the state’s driver record files, he can determine if the suspect has a prior record ofimpaired driving or other offenses. A standard interview protocol can help officers,especially inexperienced officers, acquire the information needed for trial and avoidpitfalls. Some participants have noted great success with in-car videotaping to recordthe entire sequence of stopping the vehicle and interviewing the driver at the vehiclewindow and on the roadside. Although videotaping requires additional equipment and care in operation, it can provide prosecutors and juries with direct and basicallyirrebuttable evidence of a suspect’s behavior rather than relying solely on the officer’stestimony. Once they see the taped record, offenders frequently plead guilty beforegoing to trial. Of course, tapes of drivers who appear perfectly sober may impedeconviction.


• SFST use and standardization through initial and refresher training of patrolofficers

• SFST acceptance in court• Additional states’ use of PBT at the roadside• Roadside impaired driving interview protocols• Use of in-car videotaping• Police access to driver records at roadside

IN THE POLICE STATION: BUILDING THE EVIDENCE FOR TRIAL

In the police station after arrest, the driver is interviewed and given an evidentiary breathtest. The key issue is to complete the evidence required for trial and to avoid errors thatcould cause the case to fail. Workshop participants made two suggestions. First, as at theroadside, interview protocols could help acquire the necessary information in asystematic and efficient manner. Second, some jurisdictions have found it useful tovideotape the driver’s actions while being booked, breath tested, and interviewed. Aswith roadside videotaping, many drivers will plead guilty before trial once they have seenthe visual evidence of their performance that would be presented to the jury.


• Impaired driving interview protocols for use in police stations• Greater use of videotaping


AT TRIAL: OBTAINING A CONVICTION

All the issues and evidence discussed previously come into focus if the case goes to trial.Most impaired driving cases are settled without a trial. But those cases that are triedestablish the framework within which prosecutors, defense attorneys, and judgesnegotiate settlements. Complicated statutes and procedures make it easier for cases to belost or dismissed on procedural grounds. Accurate and informed police testimony backedup by evidence from appropriate technology will help win cases. Impaired driving trialsnever will be simple, and police, prosecutors, and judges all need appropriate informationand training before they can do their part effectively.

In addition to reviewing and revising the Uniform Vehicle Code and simplifyingestablished investigation procedures, workshop participants suggested several specificchanges that could rationalize and close loopholes in the current system. In some states,impaired drivers can escape with relatively minor penalties if they refuse to take a drug oralcohol test. Should not sanctions for test refusal be at least as severe as those for takingand failing the test? In some states, drivers may request a second test using blood, breath,or urine, as they wish. This process may have been useful in the early days of breathtesting, when breath-test instruments were less reliable. Current instruments are veryaccurate, so the provision of a second test adds little except delay to the process. If stateskeep the provision, a simple way to avoid difficulty is to define impaired driving in thestate’s statutes using evidence from either blood, breath, or urine. This also avoidschallenges regarding the conversion of results from one medium to another. Some statesallow only a single test for any impairing substance. This means that if an apparentlyimpaired driver passes an alcohol test, a second test for other drugs cannot be given.Finally, some states have different laws regarding impairment for operators of boats,airplanes, and other vehicles. It is far simpler to establish uniform laws for all vehicles.

The needs are (in addition to those noted previously):

• Simplified impaired driving statutes and procedures;• Criminalized BAC test refusal;• Redefinition of impairment statutorily using either blood, breath, or urine, within

two hours of driving;• Provision for police to make more than one test for impairing substances and to

choose the tests;• Consider eliminating independent test laws;• Harmonized impaired operator laws for all vehicles;• Continued police training in techniques of gathering evidence and presenting at

trial; and• Judge and prosecutor awareness of impaired driving issues.

DRUGS OTHER THAN ALCOHOL: SPECIAL CONSIDERATIONS

Impairment by drugs other than alcohol presents a special set of challenges. Alcoholimpairment can be measured directly by the amount of alcohol in the blood or breath, andthe relations between BAC, impairment on driving-related tasks, and crash risk are verywell known. None of this is true with other drugs. There are no established relations

Hedlund B-9

between drug presence, impairment, and crash risk. Indeed, drug presence for some drugscan be detected in the body days or weeks after any impairment has ended. Consequently,state laws typically refer to impairment only due to drugs but cannot set specific illegalper se levels analogous to the 0.08 or 0.10 percent levels in state alcohol laws. Somestates have established per se laws for illicit drugs so that any presence of these drugs in adriver is illegal. Finally, the extent of driving and crash involvement by people impairedby other drugs is not accurately known at all, though it is generally believed to besubstantially smaller than alcohol-impaired driving and crashes. As a result, policetypically look first for alcohol in a driver whom they believe is impaired. If they find it,they rarely will search for other drugs. If they do not, the complexities of finding andestablishing impairment by other drugs reduce the officer’s incentive to investigatefurther.

In discussion workshop participants noted that research and clinical evidencepoint to drugs other than alcohol as a significant problem, perhaps an increasing one. Theissues presented deserve greater attention than the workshop was able to devote, giventhe issues presented by alcohol and the time available at the workshop. In the brief timeavailable for discussion, participants suggested the following specific researchdevelopment and implementation needs:

• Per se laws for illicit drugs;• Additional penalties for impairment by both alcohol and other drugs;• Observation protocols for patrol officers to recognize drug impairment at the

roadside; and• Chemical tests to identify drug use at the roadside.

CONCLUSION

The agenda described provides a large number of specific activities to improve thedetection of impaired drivers and, more broadly, to improve the entire system of impaireddriving laws, enforcement, adjudication, and sanction. But these individual actions occurwithin the overall context of the community’s policies and practices on impaired driving.As noted at the outset, the system’s goal is to deter impaired driving. Unless thecommunity supports this goal and provides both leadership and resources to carry it out,individual activities will have little effect. Furthermore, the community must decide howthese resources should best be directed: For example, what is the relative priority of adultrepeat offenders and youth covered by zero tolerance laws? Research can provide toolsfor either, but the community must decide where and to what extent these tools will beused. A summary of needs follows in the appendix.


APPENDIX: SUMMARY OF NEEDS

Overall

• Resources and priority for impaired driving law enforcement.• Public education on impaired driving.

Legal

• Simplified impaired driving laws and procedures.• Review and revision to the Uniform Vehicle Code.• Criminalized BAC test refusal.• BAC definition by blood, breath, or urine in statute.• Choice of BAC test.• “One test” laws.• Reconsider right to independent BAC test.• Test valid within defined time period.• Single DWI law for all vehicles.• Per se laws for illicit drugs.• Checkpoint use in states where legislation is required.• Use of administrative sanctions.

Program

• Training for officers (SFST, interview and trial procedures), prosecutors, judges.• Simplified DWI booking and records procedures.• Automated records systems; available online in the patrol car.• Continued development and use of technology: PAS, PBT, videotaping.• Mix of effective strategies (regular and special patrol, checkpoint).• Impaired driving patrol strategies and tactics handbook.• BAC test of all crash-involved drivers.

Research

• Better cues for use at the vehicle window (mini-HGN).• Cues for observing vehicles approaching checkpoints.• Interview techniques for roadside, crashes, and in the police station.• Cues and strategies for enforcing zero tolerance laws for youth.• Methods to acquire BAC results for injured drivers.• PAS for zero tolerance enforcement.• Procedures and effectiveness of cell phone impaired driver reporting.• Cues for identifying drug impairment at the roadside.• Improved chemical drug tests for roadside use.

Legal and Constitutional Issues Related to Detection

MICHELE FIELDS


INTRODUCTION

The objective of this paper is to provide a brief overview of constitutional and other legalissues relating to the detection of impaired drivers. For convenience sake, it is writtenwith regard to alcohol impairment, but the legal issues discussed apply to impairment byother drugs.

Within the framework of the U.S. Constitution, detection issues give rise toconcerns under the Fourth Amendment (search and seizure), the Fifth and FourteenthAmendments (due process), the Fifth Amendment (self-incrimination), and the SixthAmendment (right to counsel). Other legal issues include state constitutional law, whichmay be more protective of defendants’ rights than the federal Constitution, and statestatutory and common law, all of which vary considerably by jurisdiction and even withinjurisdictions. It is not possible within the scope of this paper fully to address each issue. Atbest, the major issues will be put into perspective, and the lesser issues will be identified.

ISSUES UNDER THE U.S. CONSTITUTION

The Fourth Amendment to the U.S. Constitution reads:

The right of the people to be secure in their persons, houses, papers, and effects,against unreasonable searches and seizures, shall not be violated, and no Warrantsshall issue, but upon probable cause, supported by Oath or affirmation, and partic-ularly describing the place to be searched, and the persons or things to be seized.

In other words, the Fourth Amendment protects individuals against unreasonablesearches and seizures conducted by government officials. “Search” and “seizure” areterms of art in constitutional law. That is, they have particularized meanings developedunder common law over the years. Not all investigatory activities are searches or seizureswithin the meaning of the Constitution. Therefore, the threshold question in FourthAmendment jurisprudence is whether the activity at issue is a search or a seizure. If not,the Fourth Amendment does not apply, and the activity is legal unless it violates otherlaws or regulations, state or federal.

Vehicle Stops

Vehicle stops by police, however brief, are seizures under the Fourth Amendment (LaFave,1996). Consequently, they must be held to be reasonable. Vehicle stops fall into twocategories: individual vehicle stops, typically conducted by an officer acting alone, and

C-1

C-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

checkpoint stops. Officers may not stop an individual vehicle merely to see if the driver isimpaired. The reasonableness element requires that there be some valid, articulable reasonfor stopping a vehicle that relates to the vehicle, the manner in which it is being driven, orthe driver’s behavior. Compelling evidence that 17 percent of drivers operating on weekendnights between 10:00 p.m. and 3 a.m. are impaired by alcohol may be sufficient to justify astop to statisticians, but it does not satisfy the Constitution (Voas et al., 1998).

Police officers deal in real-world situations that are very fluid and impossible tocatalogue comprehensively. A recent U.S. Department of Transportation study noted,“from the time an officer observes a DUI suspect (before the stop and after) everythingthat the officer observes helps him or her form an opinion . . . [on driver impairment].Every stop is different, but all of the little components add up to either arresting thedriver or letting the driver go” (Jones et al., 1998). Therefore, it is possible that somecircumstances may arise that do not fall into the following categories but which,nonetheless, will be held reasonable under the Fourth Amendment. Generally, a stopwill be upheld if

• The officer observes the driver commit an offense (a moving violation orequipment violation); or

• Erratic driving suggests driver impairment to an experienced officer; or• The officer stops the driver to render aid, offer assistance, or otherwise intervene

if he reasonably believes the person may require some assistance.

Safety belt offenses are a notable exception to the rule that police may stop anydriver observed violating a traffic law. Thirty-six states have secondary provisions intheir belt use laws that prohibit officers from stopping a vehicle solely because a driver orother passenger is not properly belted. With the exception of Minnesota, this enforcementlimitation does not apply to child restraint violations (the decision holding the Minnesotachild restraint law to be secondary is on appeal). Officers may stop drivers when theyobserve a violation of the state’s child restraint laws. There are no secondary alcohol anddriving laws. Initially, Nebraska passed a zero tolerance law for underage drivers thatcontained a provision for secondary enforcement only. The secondary provision has sincebeen repealed.

The law governing sobriety checkpoints not only instructs us about when and howcheckpoints must be conducted but also gives considerable insight into the rationale forlimiting officer discretion to make individual stops.

On November 30, 1976, at 7:20 p.m., a New Castle County, Delaware,patrolman was not answering any calls. He testified that to fill the time productively he decided to stop vehicles to conduct license and registration checks. On a relativelydeserted road, he stopped a vehicle that happened along, found marijuana in plain sightin the vehicle, and arrested the driver for illegal possession of a controlled substance.The U.S. Supreme Court ultimately decided the legality of the stop (Delaware v.Prouse, 1979). The Court held that individual officers, acting on their own initiativeand absent close supervision, may not stop vehicles without the requisite particularizedsuspicion described above. The Court was troubled by the potential for abuse that existswhen officers, unchecked by supervisory authority, are allowed to stop any vehicle theychoose. It held,

Fields C-3

except in those situations in which there is at least articulable and reasonable sus-picion that a motorist is unlicensed or that an automobile is not registered, or thateither the vehicle or an occupant is otherwise subject to seizure for violation oflaw, stopping an automobile and detaining the driver in order to check his driver’slicense and the registration of the automobile are unreasonable under the FourthAmendment (Delaware v. Prouse, 440 U.S. 663, 59 L.Ed. 2d. 67).

The Court used the term “random” to describe the patrolman’s decision to stopProuse. In fact, the lack of true, scientific randomness is what troubled the Court. Thepotential for an officer to intentionally or inadvertently introduce improper biases into theselection criteria imperils constitutional rights. In other words, vehicle stops that areconducted capriciously or arbitrarily do not withstand scrutiny under the Constitution.

Commonly, however, when the Court takes something away, it grants somethingelse. Prouse was such a case. Recognizing that the fundamental problem with thepatrolman’s behavior was that he had acted with unfettered discretion and without thechecks that the presence of other officers, supervisors, and the public would have created,the Court indicated that vehicles could be stopped without particularized suspicion, ifappropriate safeguards were established. This laid the groundwork for sobrietycheckpoints.

The seminal case on sobriety checkpoints is Michigan v. Sitz. At the trial level,the Sitz court held sobriety checkpoints invalid, finding that the arrest rate at checkpointswas not sufficiently high to justify their use under both the Michigan and the U.S.constitutions. The Michigan Court of Appeals agreed with the trial court’s interpretationof the U.S. Constitution and did not address the issue of the legality of checkpoints underthe Michigan Constitution (Sitz v. Department of State Police, 1988). The MichiganSupreme Court initially declined to hear the case. Michigan petitioned the U.S. SupremeCourt, which reversed the Michigan courts’ interpretation of the U.S. Constitution,finding that the minimal intrusion occasioned by a well-conducted checkpoint did notoutweigh the state’s interest in using sobriety checkpoints to enforce laws againstalcohol-impaired driving (Michigan v. Sitz, 1990). However, sobriety checkpoints remainillegal in Michigan. The Michigan Court of Appeals held that sobriety checkpointsviolate the Michigan Constitution (Sitz v. Michigan Department of State Police, 1992).The Michigan Supreme Court agreed, holding that Michigan police may not “engage inwarrantless and suspicionless seizures of automobiles for the purpose of enforcing thecriminal law” (Sitz v. Michigan Department of State Police, 1993).

The U.S. Supreme Court gave its imprimatur to the guidelines that the MichiganState Police used in their checkpoint. Those require supervisors to determine in advancewhen and where checkpoints are to be conducted and to be present at checkpoints. Theguidelines also establish procedures for marking checkpoints and determining whichvehicles will be stopped.

Occasionally, drivers turn around to avoid checkpoints. State laws differ withregard to whether executing a legal avoidance maneuver gives rise to reasonablesuspicion to stop the vehicle.

Since then, additional criteria have been imposed under state statutes and commonlaw. Oregon and Texas, for example, interpret the U.S. Supreme Court’s decision torequire legislative authorization for sobriety checkpoints because there was such


authority in Sitz (State v. Anderson, 1987; Garcia v. State, 1993). The extent to whichadvance publicity for checkpoints is required is a subject of debate (Commonwealth v.Amaral, 1986; Ingersoll v. Palmer, 1987; People v. Squire, 1993).

Checkpoint guidelines should contain nothing superfluous. An overly elaborateguideline may subsequently be held to have established a procedural standard that must,in all cases, be followed. Once a driver has been stopped on suspicion of alcohol-impaired driving or for some other legitimate reason, an officer has both the right and theduty to conduct a reasonable investigation. Although search issues begin to takeprecedence at this point, seizure remains an issue. The U.S. Constitution requires that aseizure be limited to no more than is necessary to accomplish the purpose for which itwas made. Thus the window of opportunity for detecting an alcohol-impaired driver in aroutine stop (checkpoint or patrol) is limited.

Roadside Investigation

As indicated earlier, a traffic stop represents a fluid, ever-changing situation. An officermust be a master of divided-attention tasks. Simultaneously, he watches out for his ownsafety and for that of the driver, vehicle occupants, and others in the area and gathers andevaluates information about the driver. Every piece of information increases or decreasesthe officer’s level of suspicion.

A point made earlier bears repeating at this juncture. All investigations are notsearches under the Fourth Amendment. Collection of impairment evidence begins oncea vehicle comes to an officer’s attention. Evidence includes the manner in which thevehicle is being operated, including the driver’s response to the request to stop. Once thedriver stops, the officer focuses attention on the driver’s appearance and demeanor. Isshe disheveled? Do her eyes focus properly and are they clear? Is her affect reasonablegiven the circumstances? Is she flushed? Can she perform simple divided-attention taskssuch as finding her license and registration while responding to simple questions? Is herspeech slurred? Is there an odor of alcohol about her or the vehicle? Has the driveradmitted to drinking? An experienced officer evaluates a driver with respect to all ofthese factors and makes an overall judgment about alcohol impairment. At the point thatthe officer determines that impairment is not likely, he has an obligation to pass thedriver through the checkpoint or, if it is a routine traffic stop, to issue the appropriatecitations, if any, and release her.

The process that has just been described is evidence collection. None of itconstitutes a search under the Fourth Amendment. A brief explanation of why theseactivities are not searches is in order.

Before 1967, courts held that a Fourth Amendment search involved a physicalintrusion into an area in which the defendant had a property interest. Katz v. U.S. (1967)greatly extended the reach of the Fourth Amendment. The case arose when investigatorsaffixed a listening device on the outside of a public telephone booth and collectedevidence of gambling offenses. The United States argued that because the defendantlacked a property interest in the telephone booth, the Fourth Amendment did not apply.The court disagreed and announced a new test.

For investigatory activity to constitute a search under the Fourth Amendment,three elements must all be satisfied. The defendant must have a reasonable expectation of

Fields C-5

privacy with regard to the evidence being sought. Society must generally recognize theexpectation of privacy as legitimate. Finally, the defendant must not have voluntarilydisplayed what he sought to protect.

Applying these criteria to the evidence gathered in the first few moments of avehicle stop, it is clear that no search occurred. The driver had no reasonable expectationthat she would not be seen or heard by a police officer legitimately stopping her on apublic road. Even if she expected to be neither seen nor heard, it is not an expectation thatsociety would recognize as reasonable. And, finally, by venturing out in public, shevoluntarily exposed herself to observation.

If the driver’s appearance and demeanor suggest alcohol impairment, the officermay incrementally escalate the investigation with a view toward establishing the presenceor absence of probable cause for an arrest. The emphasis here is on “incrementally.” Atthis point, the officer may engage the person in additional conversation. This providesfurther evidence for evaluation. The officer may ask the person to exit the vehicle toobserve the person’s coordination. Each additional piece of evidence will increase ordecrease the officer’s level of suspicion. If suspicion is not dispelled or it escalates, thenfield sobriety testing is in order.

Field sobriety testing is the final step in the investigation. The results of fieldtesting are the final pieces of evidence that determine whether probable cause for an arrestexists. Field tests fall into two general categories, behavioral and chemical. Behavioraltesting is similar to what occurs initially at the stop. Officers make close observations ofdriver behavior for signs of impairment. The difference is that the officer requests thedriver to execute very specific functions, in effect, standardized tests of coordination(walking heel to toe in a straight line, walking and turning, etc.). The officer observes andrecords the driver’s ability to focus on the instructions, understand them, and executethem. Even the gaze nystagmus test is a form of behavioral testing. The officer requires thedriver to focus his or her eyes on a particular point and observes and records involuntarymovement that is indicative of impairment.

The most common chemical test conducted in the field is the preliminary breathtest (PBT). Preliminary breath test devices are handheld devices fitted with a mouthpiece.The subject blows into the mouthpiece, providing an exclusive sample of breath. Use ofthe PBT constitutes a search under the Fourth Amendment because we recognize a strongprivacy interest in body substances taken directly from inside the body, in this case,breath from inside the mouth and deep lung air that has not been expired and commingledwith ambient air.

It is important to understand that Fourth Amendment rights may be waived.Police can always request individuals to submit to searches by consent. PBTs arerequested in this matter in several states. However, 29 states and the District of Columbiahave statutes that regulate the use of the PBT. These statutes authorize officers to requesta PBT but limit their use to circumstances in which the officer has developed reasonablesuspicion of alcohol impairment. Consequently, in these states, a PBT may not be usedearly in the investigation unless indicia of impairment by alcohol is very strong from thebeginning. PBTs are very useful because they provide scientific, objective evidence ofimpairment or violation of per se laws. However, where the law prohibits officers fromrequesting PBTs unless they have already gathered evidence of impairment, they are ofno help in the initial screening of drivers. Generally, PBT evidence is not admissible for


any purpose except to establish probable cause for the arrest. In other words, the resultsof a PBT are inadmissible in trial.

There is a critical need for an objective, reliable tool to assist officers at the verybeginning of their screening process. Such a tool must be portable, capable of being usedas soon as the stop is made; it must provide an immediate response; and its use must notconstitute a search under the Fourth Amendment. The passive alcohol sensor (PAS) isjust such a tool. The PAS is an alcohol detection device. It samples the ambient air in anarea and determines the presence or absence of alcohol in that sample. For the purposesof enforcing alcohol-impaired driving laws, the PAS should be used within 6–10 inchesof a driver’s mouth, preferably when the driver is speaking. It immediately alerts policeto the presence of alcohol in the sample.

The PAS is more objective and reliable than an officer’s nose. Field research atsobriety checkpoints shows that police commonly fail to detain 50–60 percent of driverswhose subsequent PBT results are 0.10 percent BACs. Use of the PAS significantlyimproves detection rates.

A critical issue for constitutional analysis is that the PAS is distinguishable fromPBTs and evidentiary devices by the sample it collects. Because it samples the ambient airaround a subject’s mouth, arguably in which the subject has no privacy interest, use of thePAS may not constitute a search (Fields and Henricko, 1986). As noted above, because thePBT and evidentiary tests use a sample of unadulterated breath, they clearly implicate theFourth Amendment. The question of whether use of a PAS constitutes a search has particularsignificance given the recent U.S. Supreme Court decision (Knowles v. Iowa, 1998).

Although the PAS has been used since the 1980s, there are no reported casesanalyzing the legal issues raised by the PAS under the Fourth Amendment. Furtherexplication of why use of the PAS should be constitutionally permissible is beyond thescope of this paper. The important point is that the PAS is a reliable tool that can andshould be used early in an officer’s encounter with a driver as a screening device toeliminate from suspicion drivers who have not been drinking and to focus on those whoshould be investigated for alcohol offenses. The PAS can also help an officer identifydrivers who may be impaired by drugs other than alcohol. If impairment is evident from adriver’s behavior and the PAS rules out alcohol, the officer will know to look for otherimpairing substances or for an underlying medical problem.

Probable Cause

The level of suspicion necessary for an arrest is probable cause. As one commentatornoted, probable cause is “an exceedingly difficult concept to objectify” (Cook, 1971).Probable cause is ephemeral because it is flexible. “A finding of probable cause in anygiven case rarely furnishes a formula for making similar findings in other cases becauseprobable cause depends upon the facts and circumstances of the particular case beingreviewed” (In re Armand, 1983).

Probable cause exists when known facts and circumstances are sufficient to warranta man of reasonable prudence in the belief that an offense has been or is being com-mitted . . . Probable cause . . . requires a pragmatic analysis of “everyday life onwhich reasonable and prudent men . . . act.” . . . It is to be viewed from the vantagepoint of a prudent, reasonable, cautious police officer on the scene at the time of the

Fields C-7

arrest guided by his experience and training. . . . It is “a plastic concept whose exis-tence depends on the facts and circumstances of the particular case.” It is howeverthe totality of these facts and circumstances which is the relevant consideration. . . .Viewed singly these factors may not be dispositive, yet when viewed in unison thepuzzle may fit (United States v. Davis, 1972).

The U.S. Supreme Court described probable cause as a “fluid concept—turningon the assessment of probabilities in particular factual contexts—not readily, or evenusefully, reduced to a neat set of legal rules” (Illinois v. Gates, 1983).

The officer evaluates the evidence and decides whether he has probable cause foran arrest. Although post-arrest driver behavior may provide additional evidence ofimpairment, the most significant new piece of evidence will be the evidentiary test.

Evidential Chemical Tests

Implied consent laws provide the authority for evidential testing of blood, breath,saliva, or urine. States vary with respect to the substances that may be tested, but allpermit blood and breath testing. In effect, implied consent laws establish a socialcontract, a quid pro quo. The state grants eligible drivers a license on condition thatthey consent to evidential testing when probable cause exists for an arrest for specifiedalcohol and driving violations. (In some states, arrest triggers the implied consent law.)The implied consent law is the expression of that contract. The penalty for violatingthat contract by refusing to be tested is license suspension or revocation, and somestates have actually made it a minor criminal offense to refuse to be tested whenproperly requested to do so.

Every implied consent law describes the circumstances that trigger the obligationto be tested. The officer must either have arrested the driver for the specified offenses (thestate’s equivalent of driving while intoxicated or the per se offense of driving with aprohibited blood alcohol concentration) or have probable cause for the arrest. Impliedconsent laws also are very particular with regard to the information the officer must givethe driver. All require the officer to explain the consequences of refusing the test.

Implied consent laws are the linchpin of per se laws and administrative licenserevocation laws because they create the authority for the chemical test on which the otherlaws are based. They also are highly detailed and technical. They present a trap for theunwary because if the procedures they mandate are not followed carefully, the chemicaltest may be suppressed from evidence, leaving the state to rely on behavioral evidencealone at trial.

Evidentiary tests are vulnerable in other respects as well. Both the test equipmentand the technicians who administer the tests must be certified, and the certifications mustbe current. The equipment must have been properly calibrated, and calibration recordsmaintained. Most states require that the test must be administered within a specified timeof arrest (typically two or three hours). There are detailed procedures for how the testmust be administered. If a defendant demonstrates that any of these requirements werenot met, the test may be held inadmissible.

In addition to defenses based on alleged violations of breath-testing procedures,test results have been challenged on more fundamental grounds. The scientific principle


underlying breath testing is Henry’s Law, which states that the concentration of a volatilesubstance dissolved in a liquid is directly proportional to the vapor pressure of thevolatile substance above the liquid. The lungs exchange gases between the blood and theatmosphere. As this occurs, alcohol in the blood is transferred into exhaled breath whereit can be measured. Breath-testing equipment measures the amount of alcohol in a knownamount of deep lung air and calculates from that figure the amount of alcohol in thesubject’s blood. The key is determining the proper ratio of alcohol found in the breath toalcohol found in the blood. The ratio used in every jurisdiction is 2100 to 1.

Defendants argue that the true ratio is dependent on factors like temperature andatmospheric pressure and, even if these factors are controlled, the actual ratio varieswidely among individuals and even in the same individual. Another related argument isthat the test measures the defendant’s blood alcohol concentration when administered, notwhen the defendant was driving. These arguments can turn trials into battles betweenexperts, especially in per se cases, where the offense has only two elements—operating amotor vehicle and doing so with a prohibited blood alcohol concentration.

Test evidence is used differently depending on the charge. If the defendant ischarged with operating a motor vehicle while intoxicated or under the influence of alcohol,test evidence is used in conjunction with behavioral evidence of alcohol impairment. A factfinder may determine that the behavioral evidence alone is sufficient to convict, or thatwithout regard to the ratios or other technical arguments, the breath test proves the presenceof alcohol and that, along with behavioral evidence, is sufficient. However, if the defendantis charged with a per se offense, then the test is essential for conviction.

Authority for Testing Other Than Through Implied Consent

Although most chemical tests for alcohol or other drugs are administered under impliedconsent laws, the law permits other methods of testing drivers. First, one can waive FourthAmendment rights. Drivers can voluntarily consent to testing. Second, in 1966, the U.S.Supreme Court held that involuntary testing is permissible under certain circumstances(Schmerber v. California, 1966). The driver, who was being treated in a hospital for injuriessustained in a crash, refused to submit to a blood test. Over his objections, the policedirected a physician to take a blood sample for testing. The test results were admittedagainst the driver, and he was subsequently convicted of driving while intoxicated.

On appeal, he argued that the involuntary test violated the due process protectionsof the Fourteenth Amendment, the Fifth Amendment privilege against self-incrimination,his Sixth Amendment right to counsel, and the Fourth Amendment protection fromunreasonable searches and seizures. The Court held that forcible submission to medicalpersonnel performing the minor routine task of drawing blood did not violate due process.The appropriate test for the due process analysis was whether the police action offended oursense of justice. [In an earlier case, the Court had held that forcibly pumping a suspect’sstomach was shocking and violated our sense of justice (Rochin v. California, 1952).]

Providing a blood sample did not constitute forcible self-incrimination becausethe Fifth Amendment protects persons from being required to testify. The SchmerberCourt found nothing testimonial about providing a blood sample and noted that the FifthAmendment did not prevent the state from using physical evidence. The distinction is thatthe privilege bars compelling “communications” or “testimony,” but that compulsion that

Fields C-9

makes a suspect or accused the source of “real or physical evidence” does not violate theprivilege.

Similarly, no Sixth Amendment right to counsel was violated because the purposeof the right to counsel is to assist a person to assert his other rights. As there was no right toresist testing, there was no right to counsel to protect the defendant from having to submit.

As to the Fourth Amendment claim, the Court said that it could not reasonably beargued that the Fourth Amendment did not apply to the forcible extraction of blood.However, unlike the Fifth Amendment, which prohibits compelled self-incrimination inall instances, the Fourth Amendment does not prohibit all involuntary searches andseizures, just those that are held to be unreasonable. The Court found that the police hadprobable cause to arrest the defendant for DUI. (At the crash scene, the officer smelledalcohol on the defendant’s breath and noted that his eyes were bloodshot, watery, andglassy. These symptoms of impairment were again observed at the hospital two hoursafter the crash.) Although police are permitted to search persons under arrest, the takingof a blood sample does not fall in the category of search incident to arrest because itinvolves taking a substance from within the body. The Court balanced the state’s interestin getting the test result, which is “a highly effective means of determining the degree towhich a person is under the influence of alcohol,” and the individual’s interest. Itdetermined that the police did not have to wait for a warrant because there was no way ofprotecting the evidence in the interim. The factors the Court considered in identifying theindividual’s interest included the routine nature of the test and lack of risk involved, thequalifications of the doctors who drew the blood, and the environment, a hospital, inwhich it was drawn. The Court held:

The integrity of an individual’s person is a cherished value of our society. Thatwe today hold that the Constitution does not forbid the States minor intrusionsinto an individual’s body under stringently limited conditions in no way indicatesthat it permits more substantial intrusions, or intrusions under other conditions.

Implied consent laws may have limited the application of Schmerber. For example,in Maryland, the implied consent law specifically prohibits police from compelling a personto submit to testing [MD CODE ANN., (TR) 16-205.1].

Due Process

The due process clause of the Fourteenth Amendment says that no state shall “depriveany person of life, liberty, or property, without due process of law.” The FifthAmendment also contains a similar provision.

The due process analysis requires a determination of whether the interest assertedby an individual is a life, liberty, or property interest covered by the clause. A typicalDUI/DWI conviction results in fines, possible incarceration, and license suspension orrevocation. It is obvious that the potential for a jail sentence implicates liberty interestsand that fines implicate property interests. License suspension or revocation implicatesproperty interests as well because the U.S. Supreme Court has recognized that driving isso fundamentally a part of our culture that we have a property interest in the continuedpossession of a license (Bell v. Burson, 1971).


If an interest triggers due process protection, then the relevant question becomeswhat level of process is due. This depends largely on the gravity of the individual interestat stake. Because our focus is on detecting and prosecuting alcohol-impaired-drivingoffenses, this discussion is confined to the due process concepts relevant to the criminalprosecution of DUI/DWI and not to the licensing issues.

In criminal cases, the state must prove each element of the crime beyond areasonable doubt. Consequently, irrebuttable presumptions are unconstitutional becausethey relieve the state of its burden of proof. Defendants have argued unsuccessfully thatper se statutes create such irrebuttable presumptions. Courts have reasoned that the per seoffenses have only two elements: operating the vehicle and having a prohibited alcoholconcentration while doing so. As long as defendants have the ability to test the state’scase with regard to each element, there is no irrebuttable presumption. Defendants maychallenge test accuracy on any relevant grounds (People v. Ziltz, 1983). Generally, theopinions find that defendants’ irrebuttable presumption arguments confuse DUI/DWIwith per se offenses. Because intoxication is not an element of the offense, driving at aprohibited alcohol concentration does not establish a conclusive presumption ofDUI/DWI (e.g., State v. O’Connor, 1984).

Because a driver does not know his alcohol concentration at any given time,defendants have argued that per se laws violate due process by establishing a vaguestandard of conduct. Due process considerations prohibit the state from enacting laws thatfail to put the public on adequate notice of what the exact conduct is that is prohibited.Repeatedly, courts have disposed of this argument on the grounds that a person is awareof whether or not he or she has been drinking and that a person who drives after drinkingknowingly risks violating the law (State v. Brock, 1984).

The admissibility of scientific evidence also raises due process issues. Equipmentthat purports to measure alcohol concentrations must be based on sound scientifictheories. Horizontal gaze nystagmus (HGN) evidence must be presented with theappropriate foundation, demonstrating the validity of the test and the officer’s trainingand competence to administer it (State v. Superior Court of County of Cochise, 1986;People v. Vega, 1986). Generally, HGN is more likely to be admissible to demonstrateprobable cause for an arrest than to prove intoxication or a specified alcoholconcentration at trial (Ludington, 1997).

Miranda Rights in DUI/DWI Prosecutions

No discussion of constitutional issues relating to detection of a crime would be completewithout mention of Miranda v. Arizona (1966). Miranda was designed to protect asuspect’s Fifth Amendment right not to incriminate himself and his Sixth Amendmentright to counsel. It requires the government to warn defendants who are in custody oftheir rights under the Fifth and Sixth Amendments before asking for a waiver of thoserights. Timing, therefore, is the critical Miranda issue. Miranda attaches to interrogationsthat occur after a person is in custody, not before. Also, Miranda is irrelevant if noattempt is made to interrogate an individual or to use incriminating statements theindividual may have made.

The interest Miranda seeks to protect is the right of a person not to be compelledto testify against himself. Therefore, it does not apply to non-testimonial evidence.

Fields C-11

Evidence of slurred speech is highly incriminating in a DUI/DWI case, but nottestimonial. Similarly, asking a suspect if he or she understood instructions givenregarding field sobriety or chemical tests does not constitute an interrogation that givesrise to Miranda (Pennsylvania v. Muniz, 1990).

Courts divide statements made by suspects into two categories, depending onwhat the state is trying to prove by admitting the evidence. If the content of the statementis being used against the person, the evidence is testimonial and therefore covered by theFifth Amendment. In Muniz, for example, evidence of the slurred nature of thedefendant’s speech was admissible, but his inability correctly to tell officers the year heturned six was inadmissible, because the question probed the defendant’s thoughtprocesses. HGN tests are not testimonial and cannot be suppressed because the test wasadministered prior to the Miranda warning (Smith v. State, 1986).

CONCLUSION

Alcohol-impaired driving is a criminal offense, created by statute. Consequently, all theconstitutional rights, state and federal, that apply to any criminal defendant apply to adefendant in a DUI/DWI case. That will remain true as long as DUI/DWI remains acriminal offense or until we reinterpret our Constitution. Neither decriminalizingDUI/DWI nor eviscerating the Constitution seems likely or appropriate. Whatdistinguishes DUI/DWI investigations from other investigations is the plethora ofstatutory requirements imposed on the state over and above what state and federalconstitutions demand. Other panelists will address the practical problems commonlyfaced by police, prosecutors, and judges, many of which could be eliminated or reducedby clarifying and simplifying the relevant statutes.

REFERENCES

Bell v. Burson, 402 U.S. 535, 91 S.Ct. 1586, 29 L.Ed.2d 90 (1971).

Cook, Probable Cause to Arrest, 24 Vand.L.Rev. 317 (1971).

Commonwealth v. Amaral, 398 Mass. 98, 495 N.E.2d 276 (1986).

Delaware v. Prouse, 440 U.S. 648, 99 S.Ct. 1391, 59 L.Ed.2d 660 (1979).

Fields, M., and A. R. Henricko. Passive Alcohol Sensors—Constitutional Implications. The Prosecutor,Vol. 20, No. 1, 1986, pp. 45–52.

Illinois v. Gates, 462 U.S. 213, 103 S.Ct. 2317, 76 L.Ed.2d 527 (1983).

In re Armand, 454 A.2d 1216 (R.I. 1983).

Ingersoll v. Palmer, 43 Cal.3d 1321, 241 Cal.Rptr. 42, 743 P.2d 1299 (1987).

Jones, R. K., J. H. Lacey, and C. H. Wiliszowski. Problems and Solutions in DWI Enforcement Systems.U.S. Department of Transportation, Washington, D.C., 1998.

Katz v. U.S., 389 U.S. 347, 88 S.Ct. 507, 19 L.Ed.2d 576 (1967).

Knowles v. Iowa, ___ U.S. ___, ___ S.Ct. ___, ___ L.Ed.3d ___, 1998, WL 840933 (Dec. 8, 1998).

LaFave, W. R. Search and Seizure: A Treatise on the Fourth Amendment. Third Edition. Volume IV, Sec. 10.8(a), 668. West Publishing Co., St. Paul, Minn., 1996.


Ludington, J. P. Horizontal gaze nystagmus test: Use in impaired driving prosecution, 60 ALR4th 1129(1997 Supp.).

MD CODE ANN., [TR] 16-205.1.

Michigan v. Sitz, 496 U.S. 444, 110 L.Ed.2d 412 (1990).

Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966).

Pennsylvania v. Muniz, 496 U.S. 582 (1990).

People v. Squire, 19 Cal.Rptr.2d 121 (1993).

People v. Vega, 145 Ill, App. 3d 996, 496 N.E.2d 501 (1986).

People v. Ziltz, 98 Ill 2d 38, 455 N.E.2d 70 (1983).

Rochin v. California, 342 U.S. 165, 72 S.Ct. 205, 96 L.Ed. 183 (1952).

Schmerber v. California, 384 U.S. 757, 86 S.Ct. 1826, 16 L.Ed.2d 908 (1966).

Sitz v. Department of State Police, 170 Mich. App. 433, 429 N.W.2d 180 (1988).

Sitz v. Michigan Department of State Police, 192 Mich.App. 690, 485 N.W.2d 135 (1992).

Sitz v. Michigan Department of State Police, 506 N.W.2d 209 (1993).

Smith v. State, No. 49A02-8602-CR-51 (Ind. App., 1986).

State v. Anderson, 743 P.2d 715 (1987); Garcia v. State, 853 S.W.2d 157 (1993).

State v. Brock, 357 N.W.2d 29 (1984).

State v. O’Connor, 220 NJ Super. 104, 531 A.2d 741 (1984).

State v. Superior Court of County of Cochise, 149 Ariz. 269, 718 P.2d 171 (1986).

United States v. Davis, 458 F.2d 819 (D.C. Cir. 1972).

Voas, R. B., J. K. Wells, D. C. Lestina, A. F. Williams, and M. A. Greene. Drinking and Driving in theUnited States: The 1996 National Roadside Survey. Accident Analysis and Prevention, Vol. 30, No. 2,1998, pp. 267–275.

Increasing the Opportunities to Examine Impaired Drivers

JACK W. STUSTER

Anacapa Sciences, Inc.

ABSTRACT

Measures of traffic safety have improved substantially during the past two decades.However, more than 40 percent of all fatal collisions still involve alcohol, and the risk ofbeing detected when driving while impaired remains very low, despite law enforcementefforts. This paper discusses enforcement strategies and research results intended toincrease opportunities for law enforcement officers to detect and arrest impaired drivers.Driving while impaired detection, sobriety checkpoints, and specific patrol tactics areaddressed.

INTRODUCTION

Nearly 1.4 million people have died in traffic crashes in the United States since 1966, theyear of the National Traffic and Motor Vehicle Safety Act, which led to the creation of theNational Highway Traffic Safety Administration (NHTSA) in 1970. During the late 1960sand early 1970s, more than 50,000 people lost their lives each year on our nation’s publicroads; more than half of the motorists killed had been drinking. Traffic safety has improvedconsiderably since that time: The annual death toll has declined to about 40,000, even thoughthe numbers of drivers, vehicles, and miles driven all have greatly increased. The dramaticimprovements in traffic safety are reflected in the change in fatality rate per 100 millionvehicle miles traveled: The fatality rate fell from 5.5 in 1966 to 1.7 in 1996 (Fatal AnalysisReporting System, 1996), a 69 percent improvement. Figure 1 illustrates this importanttrend. When miles traveled are considered, the likelihood of being killed in traffic in 1966was more than three times what it is today.

There have been significant improvements in traffic safety during the past 17 years;however, an average of more than 115 people still die each day from motor vehicle crashesin the United States. It is estimated that 41 percent of drivers who die in crashes havebeen drinking. Despite law enforcement efforts, the risk of detection when driving whileimpaired (DWI) is relatively low. Professor Borkenstein estimated the rate 20 years agoas one DWI1 arrest for every 2,000 trips at blood alcohol concentrations (BAC) greaterthan 0.10 percent; more recently, Ross (1992) estimated the risk to be as low as one in5,000 miles driven.

D-1

1 Various terms are used throughout the United States for offenses involving drinking and driving. In this paper, drivingwhile impaired (DWI) is used to refer to all occurrences of driving at or above the legal blood alcohol concentration(BAC) limit of a jurisdiction.

D-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

STRATEGIES FOR INCREASING DWI DETECTION

Impaired drivers are detected and DWI arrests are made as a consequence of collisioninvestigations, special events policing, traffic enforcement patrols, routine patrols, DWIpatrols, and sobriety and driver’s license checkpoints. Occasionally, law enforcementofficers are directed to impaired motorists by concerned citizens and emergency medicalpersonnel. However, most DWI arrests are made by officers on traffic enforcement,routine patrols, and special DWI patrols.

Officers who focus on DWI enforcement develop strategies to improve theefficiency of their operations. Most of the strategies are intended to increase the probabilityof encounters with impaired drivers, for example, by patrolling a course that includes barsand restaurants that are known or suspected to contribute to drinking and driving. Officershave reported in interviews that they consider it unsportsmanlike to “camp out” near aparticular establishment and wait for a patron to drive away; proprietors complain that thistactic has resulted in some establishments going out of business.

Although most officers refrain from “sitting on a bar,” they routinely make mentalnotes of vehicles parked near certain establishments. When they later observe apreviously noted vehicle departing the location, or encounter it on the road, the officersoften attempt to predict the driver’s BAC by calculating the time spent at the bar andestimating the number of drinks that were consumed. Estimates sometimes can be madewithout a confirmed duration. For example, a tradesman’s truck or van parked outside abar at 8:00 p.m. might suggest to an experienced officer that the driver has been drinkingsince leaving a work site about four hours earlier. Officers’ estimates can be uncannily

FIGURE 1 Fatality rates per million miles traveled in the United States.

Stuster D-3

accurate. Patrolling areas in which bars are located, noting the presence of vehicles, andlater stopping the vehicles for legitimate cause is a rational DWI detection strategy;devising strategies and predicting BACs on the basis of indirect evidence simply makethe job more interesting for the officers.

Most special DWI enforcement consists of roving patrols, of the type just described,in which officers concentrate their efforts on detecting and processing DWI motorists. Likefishermen who return to previously successful fishing spots, roving patrol officers typicallyfocus their efforts on areas in which DWI arrests have been made or crashes have occurred.Patrol strategies include cruising a circuit of hazardous areas and drinking locations whilelooking for vehicle code violations and other DWI detection cues. Experienced officersusually inspect the faces of oncoming drivers for the signs of alcohol impairment, inaddition to evaluating driving performance from behind. Roving DWI patrols oftenrespond to traffic collisions to determine if alcohol was involved, and they occasionallyrespond, along with other patrol units, to complaints about loud or wild parties; theseresponses are more to evaluate the driving performance of those leaving the party than toassist with noise control.

Roving DWI patrols typically are deployed at 2100 hours and relieved at 0300 hours the next morning; it has been my experience that between three and four hours are spent on patrol each night, while the remaining two to three hours areconsumed by the paperwork and other processing tasks associated with DWI arrests.Roving DWI patrols are distinguished from “saturation patrols,” which might focus onenforcement of DWI or other infractions, are geographically limited in scope, andusually are conducted by a large number of patrol units at a time. Some agenciesperiodically support special DWI patrols, such as saturation patrols, with facilitatedbooking systems (e.g., mobile command posts equipped with breath-testing equipmentand staffed by personnel to assist in the processing of arrests). Facilitated booking ofDWIs permits officers to spend more time in the field, increasing opportunities toencounter impaired drivers and to make more DWI arrests per shift.

Most DWI enforcement strategies, appropriately, focus on increasing the probabilityof contacts with adult impaired drivers. Anecdotal accounts and research evidence suggestthat adult-focused DWI strategies are incompatible with the drinking and driving patterns ofyoung drivers. Preusser et al. (1992) found that drivers under 21 years of age were involvedin fatal, alcohol-involved crashes at rates at least twice that of drivers 25 years of age andolder; young drivers, however, were underrepresented in the numbers of DWI arrests. Thedisproportionate representation of youth in alcohol-involved DWI arrest rates is attributable,in part, to a mismatch of traditional DWI enforcement strategies and the drinking and drivingpatterns of youth. Because underage drinking is illegal, it is more likely to occur at aresidence or at the beach, the river, or some other out-of-the-way place than in a bar orrestaurant. The locations where youth drink, and the routes to and from those locations, tendnot to be included in adult-focused DWI patrols. The danger associated with underagedrinking and driving is compounded by a tendency to consume all the alcoholic beverageavailable (usually because it cannot be stored), and other factors, such as an undeniable lackof driving experience and skill and youthful delusions of immortality.

An emphasis on DWI enforcement since 1980 has been a factor in the significantimprovement in traffic safety, as represented by declining fatal and alcohol-involved crashrates. NHTSA-sponsored research contributed substantially to the improved condition, in


part by providing patrol officers with useful and scientifically valid information and trainingmaterials concerning the behaviors that are most predictive of impairment. In particular,NHTSA sponsored research that led to the development of a DWI detection guide thatlisted 20 driving cues and the probabilities that a driver exhibiting a cue would have a BACof at least 0.10 percent (Harris et al., 1980; Harris, 1980). A similar study was conductedrecently that identified 24 driving cues that are predictive of DWI at the 0.08 level (Stuster,1997). NHTSA also sponsored research that led to the development of a motorcycle DWIdetection guide (Stuster, 1993). NHTSA’s DWI training materials, based on the results ofthese studies, have exposed the current generation of law enforcement officers in the UnitedStates to information critical to DWI enforcement by providing a systematic, scientificallyvalid, and defensible approach to on-the-road DWI detection.

At the same time NHTSA was providing patrol officers with informationconcerning the driving behaviors that are the most predictive of impairment, the agencyalso sponsored research that led to the development of a standardized battery of tests forofficers to administer to assess driver impairment after an enforcement stop has beenmade (Burns and Moskowitz, 1977). Beginning in 1980, officers from across the UnitedStates have been trained to detect impaired drivers on the road and to make informedDWI-arrest decisions. NHTSA has maintained an emphasis on DWI by sponsoringadditional research to develop and evaluate procedures and technologies to counterimpaired driving. The following paragraphs describe a few of the studies intended toassist officers in the detection of impaired drivers.

ON-THE-ROAD DWI DETECTION

The field detection of DWI by law enforcement officers is a problem of subtlety andcomplexity. As a consequence of observing and interpreting one or more operatorbehaviors, a patrol officer typically assesses the likelihood that a driver is DWI orotherwise impaired. This assessment then is combined with other information to reach anenforcement decision—to stop the vehicle or to continue with the patrol. Either decisionmight be incorrect. A decision to stop might result in the apprehension of a sober motorist(a false detection); a decision to permit the motorist to continue on his or her way mightresult in an undetected DWI—perhaps even a traffic collision that could have beenprevented by police intervention.

An ideal cue always would lead to a correct decision. When an ideal cue ispresent, the probability of DWI detection is 1 (a certainty); when the cue is not present,the probability of DWI detection is 0 (also a certainty). Conversely, when nighttimedrivers are tested randomly for BAC, the probability of detection (BAC = 0.08 percent)might be only about 0.04, while the probability of false detection would be 0.96. Betweenthe certainty of the hypothetical ideal cue and the probabilities of random detection, anofficer’s decision to apprehend involves the observation and interpretation of visual cuesand other information, and the subsequent trade-off between the value of a correctdetection and the cost of a false detection. The factors involved in the trade-off and thepost-detection apprehension process establish requirements and criteria for DWIdetection. In short, the detection process should employ visual cues that occur frequentlywith DWI at the lower statutory limits, are most capable of discriminating between DWIand sober operation, are simple to understand, and are easy to use by patrol officers.

Stuster D-5

Operation of a motor vehicle is a multi-dimensional task; the operator must dividehis or her attention between maintaining proper lane position and speed while monitoringthe environment for other vehicles’ movements, traffic lights, and signs. When operatorattention is divided, reaction time degrades as BAC is increased. Alcohol slows the centralprocessing of visual information; the operator’s eyes fixate for longer periods as BACincreases, apparently reducing the stimuli perceived per unit of time. This ultimately resultsin the “gazing” effect characteristic of higher BACs (Moskowitz et al., 1976; Moskowitz,1973) and contributes to performance degradation. This physiological process is translatedinto several observable driving cues. Other driving cues result from alcohol’s effect onjudgment and decision-making capabilities.

In the original NHTSA-sponsored research on DWI detection, Dr. Douglas Harrisfocused on the tasks performed by both drivers and officers on patrol (Harris et al., 1980).The focus on task performance led to a systems approach to the research, involvinginterviews with experts, analysis of DWI arrest reports, ride-along observations, BACtesting, development of prototype detection guides, and field evaluation. A key elementof this approach is the collection of driver data for all enforcement stops that are made,regardless of the dispositions of the stops. Collecting data about all observations of adriving cue permits the calculation of proportions of stops in which the cue was found inassociation with an illegal BAC level; validated proportions become the probabilities ofDWI when observed on the road. The original NHTSA-sponsored research resulted in thecreation of the DWI detection guide shown in Figure 2 and in a training booklet and film.These materials have instructed many thousands of law enforcement officers in the cues

FIGURE 2 DWI Detection Guide (Harris et al., 1980).


associated with DWI at 0.10 percent BAC and greater. The research contributedimmensely to law enforcement’s capability to counter drinking and driving by providingofficers with formal training about scientifically valid DWI cues. The training sensitizedofficers to the driving behaviors indicative of DWI. The scientific approach to theresearch, and the formal training of officers, gave credibility to officers’ statements onarrest reports and testimony in court.

Widespread acceptance of the DWI detection training materials by lawenforcement agencies and the courts prompted NHTSA to sponsor research to support thedevelopment of similar materials focusing on motorcycle DWI. The study began withinterviews of patrol officers from across the United States concerning the riding behaviorsthey had found in association with DWI. A data base containing nearly 1,000 motorcycleDWI arrest reports was assembled and analyzed, and three separate field studies wereconducted. The field studies involved the participation of 50 law enforcement sites,representing 19 separate agencies in eleven states (Stuster, 1993).

Data were collected during the field studies concerning all enforcement stopsmade of motorcyclists, regardless of the disposition of the stops. Proportion of the totalnumber of stops in which specific cues were observed in association with DWI werecalculated. A preliminary detection guide and training materials were tested during the1991 riding season in an extensive validation study. It was found that use of the detectionguide and exposure to the training materials substantially improved the abilities of lawenforcement officers to detect impaired motorcyclists, especially on the cues dependenton balance and vigilance skills. Many officers reported that before participating in thefield study, they rarely arrested motorcyclists for DWI, except at crash scenes. However,after viewing the training video, they began stopping motorcyclists for the DWI cues andmaking arrests, they believed, before crashes occurred.

A Motorcycle DWI Detection Guide, associated booklet, and 12-minute trainingvideo were developed to present the 14 rider behaviors that were found to best discriminatebetween impaired and unimpaired operation of a motorcycle. The cues are presented in twocategories. Excellent Cue Predictors (cues with a DWI probability of 50 percent or greater)are drifting during turn or curve, trouble with dismount, trouble with balance at a stop,turning problems, inattention to surroundings, inappropriate or unusual behavior, andweaving. Good Cue Predictors (cues with DWI probabilities of between 30 and 49 percent)are erratic movements while going straight, operating without lights at night, recklessness,following too closely, running a stop light or sign, evasion, and wrong way. TheMotorcycle DWI Detection Guide is presented in Figure 3.

NHTSA sponsored a third DWI detection study in 1993, in response to theincreasing number of states adopting a 0.08 percent BAC limit for DWI. The researchincluded a comprehensive review of the low BAC literature, interviews with DWI expertsfrom across the United States, development of a data base of low BAC arrest reports, andconduct of three field studies. The analysis of archival, interview, arrest report, and fielddata led to the development of a draft DWI detection guide, training booklet, and trainingvideo for evaluation; the materials included 24 driving and 10 post-stop cues.

Law enforcement agencies representing the 11 states that, at the time, had 0.08 percent BAC limits for DWI participated in the validation study. Officers reviewed thevideo and printed training materials and then completed a data collection form followingevery enforcement stop made. Data were collected during more than 12,000 enforcement

Stuster D-7

stops during this research project. The stops were made by several hundred officers,representing more than 50 law enforcement agencies from across the United States.

A feature that distinguishes this study from previous DWI detection research wasour effort to obtain low BAC data to identify any driving cues that are reliable predictors ofalcohol impairment at lower BAC levels. BACs were measured and recorded by officersduring the preliminary field study for all drivers who were found to exhibit any objectivesign of alcohol consumption (including the faint odor of an alcoholic beverage on thebreath). Although the proportions and, by extrapolation, the probabilities increase at thelower BAC levels, it is important to question to what extent the inclusion of lower BACdata contributes to a particular cue’s probability of detecting a drinking driver. In otherwords, “Are there any good predictors of low BAC levels?”

Figure 4 summarizes an analysis of the low BAC data obtained during thepreliminary field study. The seven cues listed in the figure are those that met thecriteria of having been observed at least 15 times during the preliminary study, with aprobability ( p) value of at least 0.30 at the 0.08 level. The figure shows that the lowBAC occurrences of the cues do not contribute much to the p values at the lower BAClevels, compared with the occurrences above the 0.08 level. For example, data fromthe preliminary field study indicated that the probability of finding a driver at the 0.08 level or above by stopping vehicles for all traffic infractions or behaviors was0.025, or 2.5 percent (i.e., 125 divided by 5,091). However, the probability of a BACequal to or greater than 0.08 percent was 41.4 percent if the vehicle was observedstraddling a lane line; the probability increased to 47.6 percent if the vehicle was observed

FIGURE 3 Motorcycle DWI DetectionGuide (Stuster, 1993).


weaving; and the probability jumped to 59.2 percent if the vehicle was observed to weaveacross lane lines. These cues discriminate between alcohol impairment and unimpaireddriving and provide substantial improvement in the DWI-detection capabilities of anobserver.

Similarly, it can be calculated that the probability of finding a driver with a BACgreater than zero by stopping vehicles for any traffic infraction or behavior was 0.033, or3.3 percent (i.e., 169 divided by 5,091). However, the probability of finding a driver witha BAC greater than zero but less than 0.08 percent was only 0.009, or slightly less than 1 percent (i.e., 44 divided by 5,091— 44 is the number of drivers with BACs within theselimits during the field study). Together, these data show that the low probability ofdetection at the low BAC levels contributes little to overall probability, even when cuesthat discriminate at the 0.08 level are observed. In short, the cues that are key predictorsof DWI at the 0.08 percent BAC level fail to emerge with useful p values at the lowerBAC levels (e.g., 6.3 percent for weaving).

The results of the validation study further confirmed the key cues that werecontained in the original NHTSA guide, a few additional driving cues, and 10 post-stopcues. The DWI driving cues were presented in functional categories in both the printedmaterials and the training video: Problems Maintaining Proper Lane Position, Speed andBraking Problems, Vigilance Problems, and Judgment Problems. Slight modificationswere made to the draft training materials, based on the results of the validation study. Thefinal version of the DWI detection guide is presented in Figure 5.

The new DWI detection guide and training program are distinguished from theprevious materials in several ways. One of the most important distinctions is a shift fromthe detection of “drunk” drivers, the objective of the original program, to the detection of

FIGURE 4 Contributions of low BAC casesto probabilities for key cues.

Stuster D-9

“impaired” drivers. The cues that emerged from the recent study apply to drivers withBACs of 0.08 percent and greater, and to be instructive and credible, the cues wereillustrated as realistically as possible. For example, the illustration of weaving was of thetype associated with a BAC of 0.08 percent, rather than the extreme weaving portrayed inthe original training materials. The new DWI detection materials also reflect a shift inemphasis from the probabilities of DWI to functional categories of related cues. Manyofficers reported during interviews that all cues on the guide are considered equallyimportant, regardless of differential probabilities, and, occasionally, officers’ testimonieshave been challenged when they could not recall the probabilities of specific cues. Theemphasis on categories of impaired driving behavior was intended to relieve officers ofthe requirement to memorize 34 specific probabilities. Finally, including post-stop cues inthe detection guide was intended to provide officers with additional, scientifically validinformation to help make, and later support, arrest decisions.

Several officers who were interviewed during the study mentioned that, when theydescribe a motorist’s post-stop behaviors in court, they are often challenged by defenseattorneys because information about post-stop cues usually is not included in DWI training.An officer’s extensive field experience, and a driver’s obvious signs of impairment, can beexcluded from consideration because training based on empirical data about post-stop cueshas been lacking. For this reason, 10 of the 41 post-stop behaviors that emerged from theride-along field study were recommended for inclusion in the preliminary and validationfield studies, and ultimately in the final versions of the training materials.

FIGURE 5 DWI Detection Guide developed by Stuster, 1997.


Only 10 of the post-stop cues were recommended, for a variety of reasons. Forexample, the behaviors that relate to attitude provided conflicting guidance—as manydrivers were found to be argumentative as cooperative. Furthermore, a cheerful attitudeshould not be a cause for suspicion of impairment; the implications of reasoning otherwiseare chilling. Also, cues that simply state the obvious appear to be of little possible utility toofficers (e.g., an open container). In this regard, we included the odor of alcohol from thedriver (but not from a vehicle), not because it might be useful to officers to know theobvious but to provide the basis for including the cue in formal training, which then willpermit officers to include the cue in their expert testimony.

Finally, some cues were eliminated because they might be indicators more ofsocial class than of alcohol impairment. For example, officers informed us that a flushedor red face might be an indication of a high BAC in some people. However, the cue alsois characteristic of agricultural, oil field, and other outside work. Similarly, bloodshoteyes, although associated with alcohol consumption, also is a trait of many shift workersand people who must work more than one job, as well as those afflicted by allergies. Adisheveled appearance similarly is open to subjective interpretation. We attempted tolimit the recommendations to clear and objective post-stop behaviors.

DWI DETECTION AT SOBRIETY CHECKPOINTS

The purpose of sobriety checkpoints is to deter motorists from driving while impaired.The deterrent effect of checkpoints is based on the accurate detection of DWI and publicperceptions of risk. Compton and Engle’s (1983) brief review for NHTSA was the earliesttechnical account of sobriety checkpoints (called roadblocks) as a general deterrenceapproach to DWI in the United States; they summarized the early sobriety checkpointefforts of the Delaware, Maryland, New York, and Arizona state police agencies (and a fewlocal departments) and discussed the most salient issues concerning the conduct of sobrietycheckpoints. Chief among the issues was the legality, specifically the constitutionality, ofstopping motorists without probable cause for suspicion of criminal activity. The questionof constitutionality retained its salience until June of 1990, when the U.S. Supreme Courtsettled the matter by supporting a DWI arrest made at a checkpoint by the Michigan StatePolice. But to many people, including many law enforcement officers, sobriety checkpointsrepresent an infringement of constitutional guarantees.

Several evaluations of checkpoint programs have been published since theinitial NHTSA review. Epperlein (1985) studied a brief checkpoint experiment inArizona; Mercer (1984) reported on the impact of high-visibility “roadcheck” activityin British Columbia (between 1977 and 1980); Williams and Lund (1984) evaluated theimpacts of checkpoint programs in Delaware; Voas et al. (1985) evaluated a checkpointprogram in Charlottesville, Virginia; Levy (1988) and Levy et al. (1989) evaluated acheckpoint program in New Jersey; and Lacey et al. (1990) describe checkpointprograms in Clearwater and Largo, Florida (which are combined with roving patrolscalled “wolfpacks”) and previous DWI countermeasure efforts that includedcheckpoints in Indianapolis, Indiana.

In a study evaluating screening procedures for police officers at sobriety checkpoints,cues noticed by officers were correlated with the BAC levels of the drivers. Compton (1985)found significant differences in stopping behavior. In general, drivers stopped smoothly at

Stuster D-11

low BAC levels (0.00–0.04) and “jerkily” at higher BAC levels (0.10–0.15). Drivers with alow BAC did not swerve; those with higher BACs (greater than 0.10) did swerve.

The results of the early studies showed that checkpoint programs might deterdrinking and driving, at least temporarily. However, most of the previously conductedsobriety checkpoint programs involved relatively infrequent scheduling of checkpoints.For example, in the early program evaluated by Epperlein (1985), only two checkpointsin each of three sites were conducted, and the Clearwater/Largo study conducted by Lacyet al. involved 12 checkpoints during a 15-month period. Only the programs documentedby Williams and Lund (1984) and Voas et al. (1985) were characterized by what mightbe called a vigorous program of sobriety checkpoints (i.e., 30 to 50, and 94 per year,respectively). Both of these studies found some improvement in traffic safety measuresthat were attributed to the checkpoint programs. But the studies were conducted morethan a decade ago when sobriety checkpoints were novel, extremely controversial, andnewsworthy.

In 1991, NHTSA sponsored research to systematically evaluate the absolute andrelative effectiveness of different checkpoint configurations and to compare the effects ofcheckpoint and roving patrol programs with a comparison site that received no specialtreatment (Stuster and Blowers, 1995). Six California communities participated in thestudy. Four of the communities’ police departments implemented programs of sobrietycheckpoints; the checkpoint configurations varied in terms of staffing level (three to fiveofficers versus eight to twelve) and mobility of the checkpoints (remaining in onelocation for the evening versus three sequential locations within the city). The fifthcommunity’s police department implemented a program of aggressive roving patrols thatfocused on DWI enforcement. The sixth community refrained from implementing anyspecial DWI enforcement effort for the duration of the project and served as theexperimental comparison site; statewide totals (minus the project communities) providedadditional comparison. The level of effort devoted to the roving patrols was equal to theofficer hours required to operate the high-staffing level checkpoints.

The primary dependent measure of program impact was the proportion of all injuryand fatal crashes that was alcohol-involved (BAC greater than 0.01). No significantdifferences in effectiveness of the four sobriety checkpoint programs were found. However,interrupted time series analyses found significant declines in alcohol-involved crashes in thecheckpoint communities; while the statewide totals declined, alcohol-involved crashes inthe checkpoint communities declined, on average, at a rate three-and-a-half times thecombined rate of all other communities in the state.

Further evidence of the deterrent effects of the checkpoints was found in thedeclining arrest rate during the programs. The programs began with a combined arrestrate of approximately .008, or about four DWI arrests per 500 vehicles contacted, but bythe mid-points of the programs the rate had fallen to .004, or two arrests per 500 vehicles.The arrest rate declined to .0019, or about one DWI arrest per 500 vehicles, by the end ofthe nine-month programs; that is, fewer arrests were made, despite increasing numbers ofvehicle contacts, as illustrated by Figure 6.

Officers maintained high levels of vigilance and motivation throughout thecheckpoint programs. Proficiency in detecting DWI cues in the approach lanes increasedwith experience. Officers closely monitored all approaching vehicles and even the slightestdeviation from (what officers’ considered to be) normal driving behavior immediately was


interpreted as possible evidence of a DWI. Sudden braking and lateral excursions were themost frequently observed driving cues. Officers initially interpreted attempts by motoriststo avoid checkpoints by turning before reaching the approach lane to be clear evidence ofDWI. This proved not to be the case, as many of the drivers had other reasons for wishingto avoid contact with law enforcement (e.g., illegal alien status).

Officers also scrutinized approaching vehicles for other signs of possible DWI,including furtive movements of the driver or passengers. There were many encounterswith vehicles during the 72 checkpoints conducted as part of this study in which officerswere initially certain of DWI arrests, based on the number of passengers in a vehicle,their demeanor, and the proximity of the checkpoint to well-known drinkingestablishments. Officers usually were correct in their predictions during the first fewcheckpoints but were surprised, and at first disappointed, to find with increasingfrequency, sober, designated drivers rather than the expected DWIs.

DEVELOPMENT OF AUTOMATED PROCEDURES FOR ASSESSING THEBEHAVIOR OF DRIVERS APPROACHING CHECKPOINTS

The difference between sobriety checkpoints and roving DWI patrols is analogous to thedifference between trapping and hunting strategies among commercial fishermen. Forexample, lobster fishermen, crab trappers, and most gillnetters deploy their gear inlocations known to be inhabited by the target species, in much the same way thatcheckpoints are set up at locations known for DWI arrests or alcohol-involved crashes. Incontrast, some fishermen adopt a hunting strategy by searching for indicators of fish byboth visual and technical means, then pursuing their prey, in the same manner that roving

FIGURE 6 Number of DWI arrests and vehicle contacts in four sobrietycheckpoint programs (based on data from Stuster and Blowers, 1995).

Stuster D-13

patrol officers search for, then stop, motorists who exhibit DWI cues. The trappingstrategy is fundamentally passive and dependent upon the appearance of targets in thearea; the hunting strategy is not. Hunters can increase their catch by increasing their effortwith the same amount of equipment (e.g., spending more time “on the grounds,”prospecting new areas); the means for trappers to increase their catch is to increase theamount of gear deployed. If we continue the analogy, roving patrols can increase theirDWI arrest rate by improving the efficiency of their effort (e.g., spending less timecompleting paperwork and more time on the road “prospecting” new areas for impaireddrivers, such as the out-of-the-way locations of underage drinking). In contrast, checkpointprograms rarely experience increasing arrest rates; a declining arrest rate is a measureof a checkpoint program’s deterrence on drivers. However, to be effective deterrents,checkpoints must be perceived by the public to substantially increase the probability ofdetection and arrest for those driving while impaired.

Several factors can contribute to some alcohol-impaired drivers passingundetected through sobriety checkpoints, including high officer workloads and increasingreluctance by officers to inhale the breath of hundreds of motorists. Automated DWIdetection might contribute to public perceptions of arrest risk and materially assistofficers, especially if sensors can detect subtle patterns of vehicle movements that are notreadily apparent to human observers. For these reasons, NHTSA is sponsoring research toexplore the possibility of automatically detecting impaired drivers in the approach lanesof sobriety checkpoints.

Research conducted under this contract has identified the vehicle movementscharacteristic of alcohol-impaired driving and a technology that promises the accuracynecessary to measure those movements (i.e., lateral displacement, speed and braking).The system that presently is under development has at its heart a laser speed gun,similar to those used by law enforcement agencies across the United States. Instead ofremaining focused on a target vehicle, however, the laser’s beam sweeps across thelane from left to right, then back, firing 90 times during each scan (45 times persweep). A vehicle reflects the infrared pulses when it enters the field of view of thescanning laser (beginning at a range of about 300 feet); the ranges to the vehicle andthe associated angles (from the laser to the vehicle) are obtained from these opticalreturns as the laser sweeps across the lane. Because all the ranges and angles areknown, the distance of the vehicle’s edge from the lane line can be calculated twiceeach scan. The scanning laser sweeps across the lane and back approximately threetimes each second. The prototype scanning laser has been designed to be located at theside of the road (on a low tripod); a laptop computer is connected by cable to performsystem calibration and data acquisition.

The next step in this research and development project will be to conduct acontrolled field test using dosed drivers at a simulated checkpoint. The purpose of thecontrolled field test will be to determine if drivers exhibit measurable driving behavior thatcan be correlated with BAC level. Analyses will include vehicle movement variables of(1) lateral displacement (lateral position, and the frequency and amplitude of excursions),and (2) vehicle speed (speed at acquisition, speed at points within the approach lane, andchanges in speed). Analyses will attempt to identify vehicle movements and patterns ofvariables that correlate with BACs = 0.08 percent, and if possible, lower BAC levels. In thisregard, the scanning laser system is distinguished from other efforts to detect DWI using


advanced technology by directly assessing driving performance, rather than searching forother, indirect measures of possible DWI.

The test plan has been approved by NHTSA’s Human Use Review Panel and thecontrolled field test will be conducted in October 1998. If meaningful correlations arefound during the controlled field test, the ultimate product of this research program willbe an automated system to assist officers in screening alcohol-impaired drivers at sobrietycheckpoints. The ultimate system might be slightly larger than a conventional laser speedgun and incorporate real-time processing and decision-aiding functions.

CONCLUSIONS

Operating a motor vehicle involves the performance of divided-attention tasks for whichhumans are poorly suited. A driver must attend to several sources of information, usingall sensory modalities. The driver must process the information, make a continual seriesof decisions, and translate those decisions into control adjustments, primarily involvingspeed and direction. Driving errors are common under optimum conditions, and evenminor errors and misjudgments can be fatal. The probability of error is greatly increasedwhen driver performance is impaired by alcohol.

An alcohol-impaired driver can exhibit impairment to on-the-road observers inseveral ways. Specifically, judgment can be impaired, causing a driver to follow anothervehicle too closely, make an unsafe lane change, or perform other inappropriate acts. Adriver’s ability to process information can be impaired, resulting in speed and brakingproblems, and problems associated with degraded vigilance. Impairment also is evidentin problems maintaining proper lane position, for example, weaving or turning with awide radius.

Focusing on the tasks performed by drivers on the road has permitted us toidentify 24 visual cues that are highly predictive of DWI. However, the visual cuesavailable to detect impaired drivers as they approach a sobriety checkpoint are limited bythe constrained driving conditions to vehicle speed, change in speed, vehicle lateraldisplacement, and change in lateral displacement. Because these vehicle movementvariables are among the most predictive on-the-road DWI cues, it is possible they willemerge as useful measures of driving impairment when observed unobtrusively by anautomated decision-aiding system.

REFERENCES

Burns, M. M., and H. A. Moskowitz. Psychophysical Tests for DWI Arrest. Final Report DOT HS-5-01242.NHTSA, U.S. Department of Transportation, 1977.

Compton, R. P. Pilot Test of Selected DWI (Driving while Intoxicated) Detection Procedures for Use atSobriety Checkpoints. Report DOT HS-806-724. NHTSA, U.S. Department of Transportation, 1985.

Compton, R., and R. Engle. The Use of Roadside Sobriety Checkpoints for DWI Enforcement. Report DOTHS-806-476. NHTSA, U.S. Department of Transportation, 1983.

Epperlein, T. The Use of Sobriety Checkpoints as a Deterrent: An Impact Assessment. Arizona Departmentof Public Safety, Phoenix, Ariz., 1985.

Harris, D. H. Visual Detection of Driving while Intoxicated. Human Factors, Vol. 22, No. 6, 1980, pp. 725–732.

Stuster D-15

Harris, D. H., R. A. Dick, S. M. Casey, and C. J. Jarosz. The Visual Detection of Driving while Intoxicated.Final Report DOT HS-7-1538. NHTSA, U.S. Department of Transportation, 1980.

Lacey, J., L. Marchetti, R. Stewart, P. Murphy, and R. Jones. Combining Enforcement and Information toDeter DWI: The Experience of Three Communities. Chapel Hill, NC, University of North CarolinaHighway Safety Research Center, 1990.

Levy, D. Methodologies for the Evaluation of Local Traffic Safety Programs: With an Application to NewJersey DWI Problems. Sociological Abstracts, 1988.

Levy, D., D. Shea, and P. Asch. Traffic Safety Effects of Sobriety Checkpoints and Other Local DWIPrograms in New Jersey. American Journal of Public Health, Vol. 79, No. 3, March 1989, pp. 291–293.

Mercer, G. The Relationships Among Driving While Impaired Charges, Police Visibility, and Alcohol-Related Casualty Traffic Accidents, in British Columbia, December 1977–December 1980. Ministry ofAttorney General Counter Attack Program, Police Services Branch, Vancouver, BritishColumbia,1984.

Moskowitz, H. A. Laboratory Studies of the Effects of Alcohol on Some Variables Related to Driving.Journal of Safety Research, Vol. 5, 1973, pp. 185–189.

Moskowitz, H., K. Ziedman, and S. Sharma. Visual Search Behaviors while Viewing Driving Scenes underthe Influence of Alcohol. Human Factors, Vol. 18, 1976, pp. 417–432.

Preusser, D. F., R. G. Ulmer, and C. W. Preusser. Obstacles to Enforcement of Youthful (Under 21)Impaired Driving. Final Report DOT HS-807-878. NHTSA, U.S. Department of Transportation, 1992.

Ross, H. L. Confronting Drunk Driving: Social Policy For Saving Lives. Yale University Press, NewHaven, Conn., 1992.

Stuster, J. The Detection of DWI at BACs Below 0.10. Final Report DOT HS-808-654. NHTSA, U.S.Department of Transportation, 1997.

Stuster, J. The Detection of DWI Motorcyclists. Final Report DOT HS-807-839. NHTSA, U.S. Departmentof Transportation, 1993.

Stuster, J., and P. Blowers. Experimental Evaluation of Sobriety Checkpoint Programs. Final Report DOTHS-808-287. NHTSA, U.S. Department of Transportation, 1995.

Voas, R. B., E. Rhodenier, and C. Lynn. Evaluation of the Charlottesville Checkpoint Operations. ReportDOT HS-806-989. NHTSA, U.S. Department of Transportation, 1985.

Williams, A. F., and A. K. Lund. Deterrent Effects of Roadblocks on Drinking and Driving. Traffic SafetyEvaluation Research Review, Vol. 3, 1984, No. 6, pp. 7–18.

Comments on “Increasing the Opportunities toExamine Impaired Drivers”

DOUGLAS J. BEIRNESS

Traffic Injury Research Foundation

D-16

INTRODUCTION

The detection of impaired drivers is a fundamental component of legal efforts to dealwith the drinking-driving problem. It lies at the heart of effective deterrence. As LarryRoss so aptly pointed out, effective deterrence depends on the certainty, swiftness, andseverity of punishment. By most standards, the punishments are considered severe.Administrative license suspension has undoubtedly enhanced the swiftness ofpunishment. But despite our efforts over the past 20 years, the certainty of beingapprehended and punished remains relatively low.

Years ago, Bob Borkenstein estimated that 1 out of 2,000 impaired drivers wasarrested. More recently, using data on self-reported drinking and driving combined withpolice arrest data, we estimated that the ratio was closer to 1 out of 500. Whatever theactual rate, chances are an impaired driver is not going to get caught. The bottom line isthat the probability of arrest remains low. As long as the arrest rate is low and the publicperceives the chances of arrest as low, we will not achieve true effective deterrence.

There are two choices: increase the perceived likelihood of arrest and increase theactual likelihood of arrest. There are some good examples of how the perceivedlikelihood of arrest can be increased. But unless you can fool a lot of the people for along time, the ruse will be short-lived, and the benefits temporary. Hence, this workshopdeals with strategies and techniques that will facilitate the detection of considerably morethan one impaired driver out of 2,000.

COMMENTS

Jack Stuster presents a commendable overview of the first phase involved in the detectionof impaired drivers. This can be referred to as the “active” phase. This occurs while thedrinking driver is operating the vehicle. This phase is to be distinguished from the “post-active” or “stopped vehicle” phase, which begins when the police officer has selected thevehicle from the traffic stream and the “postmortem” (or adjudication) phase, whichinvolves the examination of the evidence by the court.

There are two aspects of the “active” phase of detection discussed in the paper:(1) increasing opportunities for interacting/observing/detecting impaired drivers(increased surveillance); and (2) increasing the probability of stopping a driver who isimpaired (increased efficiency).

My comments focus on the first aspect, but not before commenting on the second.I applaud the approach NHTSA has taken in funding research to increase the efficiency ofdetection. The determination of objective cues and driving behaviors associated withimpaired driving is a laudable approach. To the extent that these cues are quantifiable and

Beirness D-17

reliable indicators of driving while impaired, they will enhance the efficiency of detectionand assist in the adjudication phase of the process. The use of such cues in the field helpsto identify impaired drivers, thereby minimizing the amount of time spent“unproductively” with non-impaired drivers.

Far too often police officers rely on inappropriate (and inefficient) cues to assist inthe detection of an impaired driver such as type of vehicle, condition of vehicle, age ofdriver, driver’s race, etc. Intermittent reinforcement perpetuates the use of suchindicators. Unfortunately, not only is it difficult to justify the use of such cues in court butit also results in officers missing large segments of the impaired driver population. Themore objective the cues, the better.

Before police officers can use any of the detection techniques outlined by JackStuster, they must first be in a position to observe driver behavior and interact withdrivers. This is done in a variety of ways, including routine patrols, collisioninvestigations, special traffic enforcement campaigns (e.g., seat belts, license checks),citizen reports, special DUI patrols (e.g., saturation patrols), and sobriety checkpoints.

Each method serves a unique function and probably targets a slightly differentsegment of the driving while impaired population. Hence, reliance on a single techniqueis ill-advised.

Improved detection involves more than simply increasing enforcement. As Stusterpoints out, the key is to implement better, more efficient techniques.

Over the past several years, a great deal of attention has been directed at sobrietycheckpoints. Special enforcement efforts of this nature are personnel- and cost-intensive.Although checkpoints create the opportunity for intersecting with a large number ofdrivers (which is important for general deterrence), they tend to find a relatively smallnumber of impaired drivers. From this perspective, they are inefficient. This is not tosuggest, however, that they are ineffective. There are several examples of highly effectivesobriety checkpoint campaigns. Rather, the checkpoints are an inefficient means ofdetecting impaired drivers. Perhaps the greatest value of sobriety checkpoints is that theyput the officers out there on the road actively looking for impaired drivers. In addition tothe public relations value of this exercise, it serves to increase the public’s perception ofthe likelihood of arrest.

The inefficiency of sobriety checkpoints is illustrated by the discrepancy betweenthe proportion of impaired drivers identified through roadside surveys and the proportionarrested at checkpoints. We know that police, even when face to face with drivers atsobriety checkpoints, fail to detect more than half of all impaired drivers. This situation is ofconsiderable concern. Not only does it underscore the difficulty of the task of detection butit also highlights the need for more effective detection techniques. In addition, the failure todetect impaired drivers at checkpoints only reinforces the behaviors of those drivers whomanage to escape detection. If the police do not suspect these drivers of being impaired andallow them to proceed, they are unlikely to be deterred from engaging in the same behavioron subsequent occasions. It is, therefore, imperative that we undertake further research tounderstand who these drivers are and why they are not detected.

Jack Stuster also alludes to the fact that not all checkpoints are created equal.They differ in the number of officers and vehicles employed, the number of locations pershift, the sites used, the time of year, etc. I do not believe we know which approach orcombination is most effective. In fact, there are often disagreements concerning whether


checkpoints should be set up early in the evening so as to interact with as many drivers aspossible, or later in the evening when the probability of encountering drunks is greater.Further information would facilitate the implementation of the most effective andefficient type of checkpoints.

In his paper, Stuster indicates that officers sometimes try to make their jobs moreinteresting by attempting to guess the BAC of a suspected offender. The issue, however,goes beyond simply making the job interesting. The success of any particular enforcementtechnique or detection strategy ultimately depends upon officer motivation and training.Officers involved in DUI enforcement must receive encouragement, support, reinforcement,and DUI arrests. DUI surveillance is an important aspect of police work and should notbe perceived as second-class duty. It is the responsibility of senior police officers, theadministration, and the public to ensure that officers engaged in DUI enforcement receivethe same support, credit, and encouragement as officers doing other types of police work.

In this context, the “system” often works against us. A simple DUI arrest can takeseveral hours to process. All too often an officer’s time and effort are for naught becausethe courts rule in favor of the accused. This can be most discouraging and affects themorale of all officers engaged in DUI enforcement. Moreover, it creates a situationwhereby officers may exercise discretion and use other means to remove potential DUIoffenders from the road without laying formal charges. Greater efforts must be made tosimplify the procedures and necessary paperwork to lessen the burden on officers and toreduce the opportunity for errors that can result in a case being dismissed on a technicality.A simplified process would also allow officers to spend more time on the road.

As a final point, I would like to at least raise the possibility of pushing thedetection process further upstream to include a “preactive” phase—that is, before thedrunk gets behind the wheel. While this issue may well be beyond the scope of thisworkshop, I think it would be a mistake if we restricted our view of detection tosituations in which the impaired individual is actively posing a risk to all road users.While Stuster indicates that officers may view it as “unsportsmanlike” to lie in waitoutside establishments known as sources of impaired drivers, it is not inconceivable forthem to enter licensed premises on occasion to remind potential drivers and/ormanagement that the police will be looking for impaired drivers. Enhanced serverintervention programs might also be implemented. Finally, we might also re-visit thedevelopment of reliable and valid “self-detection” techniques to assist drinkers and/ortheir companions in determining their fitness to drive.


BARRY M. SWEEDLER

National Transportation Safety Board

D-19

The paper by Jack Stuster does a very good job of outlining the enforcement strategiesand the research results intended to increase opportunities for law enforcement officers todetect and arrest impaired drivers. It focuses on DWI detection, sobriety checkpoints, andspecific patrol tactics. Dr. Stuster also discusses new technology to detect impaireddrivers at checkpoints.

I would like to expand on some of Dr. Stuster’s points and discuss a number ofadditional strategies that I believe will increase the opportunities that law enforcementofficers have to come in contact with impaired drivers.

FACILITATING BOOKING OF DWIs

Dr. Stuster notes that some agencies support special DWI patrols with facilitated bookingof DWIs. This can take the form of mobile breath-testing vehicles and personnel to assistin processing the arrests. This is a subject that has been discussed for more than twodecades. Officers still complain that it takes many hours to process a DWI arrest. Thetime has come to stop talking about the problem and do something about it. A report orset of guidelines to assist the arresting officer in significantly reducing the time spentoff the road is overdue. The guidelines should include suggested methods forcomputerizing the reports that an arresting officer must complete. In this computer age,there is no excuse for an officer to be required to provide handwritten copies of the sameinformation on many forms. Getting the arresting officer back on the road quickly is onesure way to increase an officer’s contacts with impaired drivers.

ENFORCEMENT OF YOUTH DWI AND ZERO TOLERANCE

In his paper, Dr. Stuster reviews the cues that have been developed to detect impaireddrivers and motorcycle drivers and for post-stop procedures. He also cites Dave Preusser’s(1992) work on detecting youth DWI. Preusser et al. found that young drivers aresignificantly underrepresented in DWI arrests compared with fatalities and have differentdrinking and driving habits. They drink at different times and places than older drivers anddo not display the same impairment cues as adult drivers. The arrest process for juvenilesis also more difficult and time-consuming. A number of successful programs to detect andprevent youth drinking and driving have been documented (Center for Substance AbusePrevention, 1995). These include special patrols at times and places when and whereyoung drivers drink, special training of officers, encouragement from high-rankingofficials, and techniques to prevent the young impaired drivers from scattering. However,these programs could be more widely adopted. There is also evidence emerging that thezero tolerance laws that are now in effect in all states are not being well enforced (Voas,1998). Based on this discussion, the law enforcement community needs specific


guidance or cues to detect young drinking drivers, information on where and whenthey drink and drive, and details on programs that are effective.

REPORT EVERY DRUNK DRIVER IMMEDIATELY PROGRAMS (REDDI)

One way to increase police contacts with impaired drivers is to turn every driver into apotential reporter. In the 1980s most states adopted REDDI programs that encouragedmotorists who observed certain erratic behaviors of other drivers to call and report thatinformation. The programs were well advertised, using roadside billboards. Thoseprograms required motorists to drive to a telephone or use a CB radio to make the report.Even with the time-delay problems, thousands of impaired drivers have been detected inthis manner (National Transportation Safety Board, 1982). Many of these programs arestill in operation but are not being publicized and have different telephone numbers tocall. Now that many drivers have cellular telephones, these programs, if well publicized,can have an ever greater impact on both increasing police contacts with impaired driversand deterring their behaviors. A new emphasis, perhaps in conjunction with thecellular telephone industry, is needed to revitalize this important program.

PERSISTENT DRINKING DRIVER

While Dr. Stuster does not mention any special enforcement campaigns aimedspecifically at the persistent or repeat offender drinking driver, a number of states andcommunities have implemented special emphasis programs to improve their chances ofcatching these multiple offenders, who in most instances should not be driving. Specialprograms aimed at apprehending persistent and repeat offenders should be greatlyexpanded. In 1991, Ohio, as part of its Habitual Offenders Program, instituted theHabitual Offender Tally, or HOT sheet, which lists offenders who have been convicted ofDWI five or more times and whose driving privileges are currently suspended. The list isshared with state and local police. Special awards recognize officers who arrest those onthe HOT sheet. Another program, known as Stakeout, is in use in some states, includingNew York. Police officers watch the homes of repeat offenders who have lost theirlicenses, usually at times when people would leave or return from work, to see if theydrive. Special license plates, such as the Zebra tags in use in Oregon and WashingtonState, have proved useful for officers to make contacts with repeat offenders.

SOBRIETY CHECKPOINTS

There is quite a bit of discussion on sobriety checkpoints in Dr. Stuster’s paper. Iwould like to mention a few additional items that should be considered when sobrietycheckpoints are planned and conducted. One important step at all checkpoints shouldbe checking for a valid license. There also should be greater use of passive alcoholsensors. The lessons learned from the random breath-testing programs in Australia andother countries, and from the successful programs of Booze It & Lose It in NorthCarolina and Checkpoint Tennessee, should be reviewed carefully. Guidelines on themost effective way to conduct sobriety checkpoints should be developed for use bystate and local law enforcement agencies.

Sweedler D-21

REFERENCES

Center for Substance Abuse Prevention. Law Enforcement Action Guide, Teen Drinking PreventionProgram. DHHS No. 95-3028. Washington, D.C., 1995.

National Transportation Safety Board. Letter to the States Recommending the Implementation of CitizenReporting REDDI-type programs. NTSB, U.S. Department of Transportation, Sept. 9, 1982.

Preusser, D. F., R. G. Ulmer, and C. W. Preusser, Obstacles to Enforcement of Youthful (Under 21)Impaired Driving. Final Report DOT HS-807-878. NHTSA, U.S. Department of Transportation, 1992.

Sweedler, B. (Ed.). Strategies for Dealing with the Persistent Drinking Driver. Transportation ResearchCircular 437, TRB, National Research Council, Washington, D.C., 1995.

Voas, R. B. Presented at the Alcohol Policy XI Conference, American Medical Association, Chicago, Ill.,1998.


JAMES C. FELL

National Highway Traffic Safety Administration

D-22

ABSTRACT

More people have been arrested for driving while intoxicated (DWI) over the past 20 years than for any other crime in the United States. Yet it is estimated that only onearrest is made for every 500 to 1,000 impaired driving trips. One way to increase DWIdetection, to generally and specifically deter impaired driving, and to reduce impaireddriving fatal crashes is to institute highly publicized, frequent sobriety checkpointsusing passive alcohol sensors. With the increasing use of seat belt, driver’s license,sobriety, and other safety checkpoints, passive alcohol sensors can increase DWIdetection dramatically. Cost effectiveness of this approach must be better communicatedto the enforcement community. Methods to promote the use of checkpoints must bedeveloped.

Although it is true that nearly 1.4 million people have died in traffic crashes in theUnited States since 1966, we arrest 1.4 million drivers each year for driving whileintoxicated (DWI) or driving under the influence (DUI) of alcohol. More people havebeen arrested for DWI in the past 20 years than for any other crime (Federal Bureau ofInvestigation, 1997). If self-reported drinking and driving is accurate, there werealmost 800 million episodes of drinking and driving in 1995. Assuming most of thosepeople were impaired, that means we arrested only 1 driver out 556 drinking anddriving trips in 1995 (Jones and Boyle, 1996). This is probably a low estimate, but wewill find out how low when work is completed by one of our colleagues here at theworkshop, Paul Zador.

Dr. Jack Stuster’s paper on “Increasing the Opportunities to Examine ImpairedDrivers” is a thorough review of the strategies to increase DWI detection. He mentionsthat impaired drivers are detected and DWI arrests made as a consequence of collisioninvestigations, routine patrols, sobriety checkpoints, etc. He does not mention an ever-increasing enforcement measure that will certainly play a role over the next few years—seat belt usage enforcement checkpoints. These have great potential indetecting impaired driving as evidenced by the “Click It or Ticket” campaign in NorthCarolina (2,000 arrests for DWI in addition to 59,000 citations for restraint usageviolators).

Roving patrols and saturation patrols may be the most popular form of policedetection of DWI, but they may not be the most effective. There is growing evidence, asDr. Stuster points out later, that sobriety checkpoints may be the most effective method ofdeterring DWI. Random sobriety checks, especially when they are conducted inconjunction with the use of passive alcohol sensors (PAS), have the potential to

Fell D-23

(1) Detect more impaired drivers,(2) Serve as a general deterrent to drinking and driving, and(3) Serve as a specific deterrent to impaired driving, if enough drivers are caught.

Dr. Stuster has a good discussion of sobriety checkpoints in his paper, but hefailed to reference one of the recent and most important studies to date—the effectivenessof the “Checkpoint Tennessee” program (Lacey et al., 1997). That study showed thatwhen a sobriety checkpoint program is conducted every weekend, all around the state, atrandomly selected sites, is highly publicized, and uses passive alcohol sensors to helpdetect DWI, a 22 percent reduction in the number of drunk-driving related fatal crashes (a reduction of nine per month) can be expected. Drinking and driving went down inTennessee and the reported use of designated drivers went up (as noted by the police atthe checkpoints).

Dr. Stuster also gives a thorough review of the cues police can use to detectdrivers at BACs greater than or equal to 0.08 g/dl. It is interesting to note that “Speeding”or “Going 10 Mph Over the Speed Limit” are not in the top 20 cues. For those who thinkimpaired driving always results in speeding or other reckless driving, this is importantinformation. “Driving Too Slow” is in the top 10 cues, but “Speeding” is not. Furtherresearch, however, could show that young drivers (under age 21) with low BACs mightexhibit that “speeding” behavior more often than adult drivers.

Finally, Dr. Stuster is conducting important research on automated proceduresfor assessing the behavior of drivers approaching sobriety checkpoints. Whatever theresults, it is strongly recommended that the costs of such a detection system areconsidered and compared with the cost of each police officer using a passive alcoholsensor. Passive sensors have a higher detection rate of impaired driving than any otherprocedure to date. They may be considered costly and burdensome by some in thepolice community, but they have great potential for DWI detection (with very littletraining required).

In conclusion, considering the low DWI detection rate in the United States, we must find better procedures and ways to increase that rate. Dr. Stuster gives acompelling description of the cues that police should look for to detect DWIs. Whilethese cues will help the police, a bigger view of the problem dictates procedures that will ultimately deter impaired driving, which, it is hoped, is the goal of anyenforcement program. Highly publicized, frequently conducted checkpoints (whetherthey are for seat belt usage, for driver’s license checks, or for detection of impaireddrivers) will serve as both a general and a specific deterrent to DWI. Passive alcoholsensors will increase detection at these checkpoints. If properly conducted, even highBAC-experienced drunk drivers could be specifically deterred because they know aPAS will detect them as drinking drivers when other behaviors they have learned tomask may escape police detection. Passive alcohol sensors will also detect commercialdrivers at illegal BAC levels (0.04 BAC) and under age 21 drivers at low BACs toenforce “Zero Tolerance” laws. While cost and police resources are important issues in this approach, it nevertheless appears to be cost effective based upon the research.Better ways to promote checkpoint/passive sensor use must be developed and barriersto their use must be overcome.


REFERENCES

Federal Bureau of Investigation. Crime in the United States, 1996. Uniform Crime Reports. U.S. Department of Justice, Washington, D.C., Sept. 28, 1997.

Jones T., and J. Boyle. National Survey of Drinking and Driving Attitudes and Behavior, 1995. Schulman,Ronca and Bucuvalas, Inc. Final Report DOT HS-808-438. NHTSA, U.S. Department ofTransportation, July 1996.

Lacey, J. H., R. K. Jones, and J. C. Fell. The Effectiveness of the “Checkpoint Tennessee” Program. InProceedings of the 14th International Conference on Alcohol, Drugs, and Traffic Safety, Vol. 2, 1997,pp. 969–975. CERMT, Annecy, France.

Identification of Alcohol Impairment on Initial Interview

DAVID F. PREUSSER

Preusser Research Group, Inc.

Earlier papers described overall impaired driving issues, legal issues, and stoppingvehicles from a moving traffic stream. After a vehicle is stopped, the next step in theprocess is the driver interview, typically conducted by the officer at the driver-sidewindow of the vehicle. The purpose of this interview, in conjunction with otherobservations made by the officer, is to determine whether or not there is reasonablesuspicion to proceed with an impaired driving investigation. If yes, the driver is typicallyasked to step out of the vehicle and submit to field sobriety testing. If no, no furtherimpaired driving investigation would likely be conducted.

The objective of this paper is to summarize available research on process andprocedure as they relate to an impaired driving investigation at the driver-side windowimmediately following a vehicle stop. Throughout this paper, the term DUI encompassesdriving while intoxicated; driving under the influence; operating while intoxicated;operating under the influence; operating at a BAC above 0.08 or 0.10 percent, and similarcharges. It should be noted that this general use of the term DUI obscures importantdistinctions between each charge as defined uniquely in the laws of each state.

BACKGROUND

Enforcement of impaired driving laws is conducted by police officers who are membersof municipal departments, state police agencies, highway patrols, sheriff’s departments,and a variety of other police agencies. Collectively, these agencies have approximately600,000 sworn personnel and make nearly 1.5 million DUI arrests each year (FBI, 1997).Nationally, the number of arrests has declined over the past 10 years.

Many of these arrests are being made by officers assigned to police entities whoseprimary mission is traffic. These officers may be part of the highway patrol, other statepolice organizations dedicated to traffic, the traffic division of a municipal or county-level department, or a dedicated DUI patrol. Such officers make up only a portion of thetotal complement of sworn personnel. Regular or precinct patrol officers may or may notbe conducting impaired driving enforcement depending on their training, theirdepartment’s emphasis on impaired driving enforcement, and the demands placed uponthem for other types of police services.

A DUI arrest can result from regular patrol activity, a crash investigation, or somespecial operation such as a sobriety checkpoint. Patrol activity, including DUI patrols,provides the greatest number of arrests, followed by crash investigations and specialoperations. Patrol activity provides the officer with the greatest amount of information priorto the driver-side interview; sobriety checkpoints provide the least amount of information.

For patrols, DUI detection relies on officer experience, stopping many motoristsfor observed moving violations, and/or a set of DUI detection cues (Stuster, 1997).Typically, the officer has observed the on-road driving behavior of the motorist before

E-1

E-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

stopping the vehicle and commencing the driver-side interview. In fact, the officer mayhave followed the motorist for some time prior to the stop and already suspect that animpaired driving violation has occurred. Also, typically, the officer can concentrateentirely on this motorist rather than on the many concerns present at a crash scene orother vehicles waiting to go through a checkpoint.

At a crash scene, the first priority is to secure the location and provide assistanceto those who may be injured. Most often, the officer has not seen on-road drivingbehavior, only its consequences. It may be difficult for the officer to pursue a DUIinvestigation when the suspected driver has been injured. People may be reluctant toprovide evidence since a DUI arrest following a crash could have both criminal and civilliability implications. An actual driver interview may never occur or may occur severaldays later in the hospital. In the 1960s, it was not uncommon to find departments thatrecorded 50 percent or more of their DUI arrests at crash scenes, and some such agenciespersist today. However, as DUI arrests from patrol activities have increased, it is moreusual now to find only one-quarter or fewer DUI arrests being made following crashes.

Checkpoints offer a third, characteristically different set of circumstances. First,unlike observations during patrol, at a checkpoint the officer has only a very limitedopportunity to observe on-road driving behavior. The driver-side interview can be virtuallythe only source of information for determining whether or not to proceed with an impaireddriving investigation. Second, a police agency’s authority to conduct checkpoints is oftenrestricted by law, regulation, or agency policy to ensure that motorists move through thecheckpoint with a minimum of delay. Such restrictions can have the effect of placing timeor procedural requirements on the officer that are not necessarily present during normalpatrol. For instance, at some checkpoints, officers do not request the license andregistration of all motorists entering the checkpoint. Yet, a driver’s ability to produce alicense and registration without fumbling is an important cue for identifying impairment(Stuster, 1997). Available research indicates that more than half of all drinking drivers aremissed at checkpoints (Jones and Lund, 1986; Ferguson et al., 1995; Wells et al., 1997).

It should also be noted that there are differences in the characteristics of drinkingdrivers arrested during the various types of enforcement activities. For example, 23 percentof drivers arrested at Charlottesville, Virginia, checkpoints were under the age of 21 ascompared with only 11 percent of drivers arrested by patrol activity during the same period(Voas et al., 1985). In Connecticut during 1997, 11,747 drivers were arrested for DUI (dataprovided by Connecticut DOT). Of these, 23 percent were arrested as a result of a motorvehicle crash while most of the remainder were arrested from patrol operations. The crasharrests were significantly more likely to occur on Mondays through Thursdays; duringdaylight and early evening hours; involve both more younger and more older drivers; morefemales; and much more often involve very high BACs (41 percent at 0.20 percent BAC orgreater versus 30 percent at that level for other DUI arrests).

OBSERVABLE CUES

As discussed earlier by Fields, an officer begins the driver interview with some level ofsuspicion regarding impaired driving. This level might be near zero for, say, a routinetraffic stop on a Tuesday afternoon and very high if the officer has followed the vehiclefor some time on Friday night and observed multiple DUI on-road driving cues. Evidence

Preusser E-3

collected during the interview will lower this level of suspicion or raise it to the point ofrequiring the motorist to exit the vehicle for field sobriety tests. Such evidence mayconsist of observable cues, described in this section, or chemical tests, described later.

An extremely important post-stop cue that might be observed during the driverinterview is an open container, drugs, or drug paraphernalia. The officer may also obtaina direct admission of the use of alcohol or some other drug. The importance of this typeof information, when it is present or if it can be developed, should not be underestimated.

Ulmer et al. (in process) obtained results for 1,125 non-alcohol drug impaireddriving suspects in Colorado during the period 1994–95. Of these suspects, 58 percentadmitted drug use at some point during the arrest process; 27 percent denied drug use; anddrug use admission or denial was unknown for the remaining 15 percent. In cases where thesuspect admitted drug use, the officer’s opinion regarding which drugs had been used wasconfirmed by laboratory analysis 93 percent of the time. In cases where the suspect denieddrug use, the officer’s opinion that drugs were involved was confirmed only 55 percent ofthe time. Similarly, for 341 Arizona cases during 1995, admitted drug use was confirmed bythe laboratory 90 percent of the time, while for denials and drug use admission unknown,laboratory confirmation of an officer’s opinion that drugs were present was confirmed only69 percent of the time. Another observable post-stop cue that can be obtained as part of thedriver interview is the smell or odor of alcohol. Moskowitz et al. (1997) dosed subjects to0.00 percent BAC, at or below 0.08 percent BAC, and above 0.08 percent BAC. Samplesof exhaled breath from each of the subjects were delivered to police officers across anopaque screen through a plastic tube two inches in diameter. Results for these “optimum”laboratory conditions indicated 78.5 percent correct police officer judgments. Compton(1985) tested dosed subjects at “simulated” checkpoints. Police officers reported the odorof alcohol 61 percent of the time for subjects at BACs of 0.10 percent to 0.15 percent and39 percent of the time for subjects at BACs of 0.05 to 0.09 percent. This compares with 7 percent of the time for subjects at BACs of 0.00 to 0.04 percent. Clearly, the odor ofalcohol is an extremely important post-stop cue.

Stuster (1997) also recommended that the odor of alcohol should be used,along with several other post-stop cues. To be recommended, a post-stop cue must beobserved often enough to be useful and, when observed, must be correlated withimpairment. The cues, total number of times observed, number of times associated with a DUI arrest, and percent of observations associated with a DUI arrest are listed in Table 1. These data cover 4,604 total vehicle stops, made by 11 police agencies,resulting in 574 DUI arrests.

TABLE 1 Post-Stop Cue Data


Some officers introduce “divided attention” as part of the interview process.Typically, the officer will ask some question or questions while the driver is looking forhis license and registration. Compton (1985) found only a modest relationship betweenmultiple divided-attention types of cues and driver BAC at “simulated” checkpoints.

Horizontal gaze nystagmus (HGN), a key component of the Standardized FieldSobriety Test, is being used by some officers as a screening test part of the driver interviewwhile the driver is still seated in the vehicle. An officer asks the suspect to focus his eyes onan object, such as the tip of a pen, while that object is moved far to the right and far to theleft. HGN refers to the involuntary jerking of the eyes as they are turned as far as possibleto the side. Impaired drivers are often unable to follow the slowly moving object smoothly;exhibit nystagmus sooner at a more shallow angle; and exhibit greater nystagmus when theeye is as far to the side as possible (Tharp et al., 1981).

Compton (1985) found that HGN could be “easily and accurately administered”to a driver seated in the vehicle at a simulated checkpoint. Police officers correctlyidentified 95 percent of the drivers at BACs of 0.10 percent or greater while incorrectlyidentifying only 15 percent of the sober drivers. More recently (McKnight et al., inprocess), it has been confirmed that HGN can be administered to a driver seated in thevehicle. In this same study, HGN was the “most valid indicator” of BACs above andbelow 0.04 percent and was the only one of the three tests in the Standardized FieldSobriety Test Battery that distinguished BACs in the 0.08 to 0.10 percent range.

In summary, as part of the driver-side interview, the officer may obtain anadmission of drinking, or of taking some other drug, or may observe direct evidenceindicative of alcohol or other drug use. The officer may also detect the odor of alcohol.Next, there are a series of cues that, when observed, are indicative of impairment. Theseinclude slurred speech, psychomotor failures or decrements, the suspect’s providingincorrect information, and the suspect’s needing extra time or extra instruction for simplecognitive tasks. HGN can be administered at the driver-side window and can be a veryeffective screening test should police agencies choose to use it in this way. Currently, it isoften the case that drivers at or above 0.10 percent BAC are not detained at checkpoints (55percent, Jones and Lund, 1986; 45 percent, Ferguson et al., 1995) or by officers on normalpatrol (31 percent, Kiger et al., 1991) or on special DUI patrol (12 percent and 44 percent,Lund and Jones, 1987). Perhaps observable cues could be aided by chemical tests.

CHEMICAL TESTING

In some countries it is possible to conduct a direct active chemical test to screen drivers atthe roadside with or without probable cause for a DUI arrest. This is not true in theUnited States. A direct chemical test of a bodily substance, such as blowing into a breathtest device, is conducted only after an arrest is made or only on a voluntary and non-evidentiary basis with the consent of the suspect.

Still, chemical tests have the potential of providing the officer with an efficientand objective measure of alcohol. One technology that has been used to obtain such testsis the passive alcohol sensor (PAS). The intent of these devices is to capture and test ameasured amount of exhaled breath of the driver during the driver-side interview.Currently available devices must be held very close to the driver’s mouth and areactivated just as the driver is speaking. The exhaled breath is considered to be “in plain

Preusser E-5

view” and thus can be captured and chemically tested without probable cause for a DUIarrest. The test is passive in that it does not require the active cooperation of the driverother than answering the officer’s questions during the interview. PAS devices have beenshown to be particularly useful at sobriety checkpoints, where the officer has littleinformation about the driver prior to initiating the driver-side interview.

PAS devices of the earliest design were tested by the District of Columbia patrolofficers. The results indicated that drivers above the 0.10 percent BAC legal limit couldbe reliably identified (Voas, 1983). Second- and third-generation PAS devices have beentested at checkpoints. Results indicate that officers who are using these devices, versusofficers who are not using the devices, are able to detain more drivers at BACs of 0.05 to0.09 percent and 0.10 percent and higher. These same officers are also able to let passmore drivers at BACs of 0.04 percent or lower (Jones and Lund, 1986; Ferguson et al.,1995; see also Preusser, 1994).

It has also been found that PAS screening devices can improve alcohol detection aspart of normal patrol operations (Kiger et al., 1991). However, the currently available PASdevices are not well accepted by patrol officers (Leaf and Preusser, 1996). Such devicesforce the officer to be in closer proximity to the driver than is normal on a routine patrolstop. They require the officer to devote some level of concentration to the device at a timewhen the officer has several other things to be concerned with, including personal safety,being alone at night with an unknown driver and possibly passenger(s). Some PAS devicesrequire that the officer devote one hand to their operation at a time when that hand might bebetter left free to reach for a weapon, should that become necessary.

Officer safety, security, and proximity to the driver are of somewhat less concernat sobriety checkpoints than on normal patrol. Checkpoint locations typically have goodlighting and are well secured, and each officer has the support of several other officersduring the operation. Also, the information provided by PAS devices is seen to be ofgreater value at checkpoints since the officer has had little opportunity to view on-roaddriving behavior.

Another potential technology, under development, is the use of a laser beam todetect alcohol. The beam is passed through the passenger compartment windows. Ifalcohol is present in the air of the compartment, then the beam will be modified in waysthat can be detected. It is not known whether the practical effectiveness of thistechnology will eventually be demonstrated (Fell, 1998). In summary, direct chemicaltesting requiring the active cooperation of the driver is not conducted during the impaireddriving investigation that occurs at the driver-side window. Still, PAS devices can beused to sample the exhaled breath of the driver. Currently available devices have beenshown to improve officer efficiency at a checkpoint, both in terms of detaining moredrivers who had been drinking and of not detaining drivers who have had no drinks oronly one or two. PAS devices are not generally used for vehicle stops made while onpatrol. Other technologies are currently under development.

FUTURE RESEARCH

The following are research topics that could be pursued in the future. The topicspresented are not all-inclusive nor are they presented in sufficient detail for work tobegin. Rather, the list is intended as a starting point for a more comprehensive discussion.


1. Persons missed as part of impaired driving screening at checkpoints are not arandom sample of all drinking drivers at the checkpoints (see, e.g., Wells et al., 1997).Are there other cues, or would better use of current cues or some other procedures ensuredetaining young people, women, and other persons who are more often missed?

2. Should HGN be used for initial screening while the driver is still seated in thevehicle?

3. Can we develop procedures for establishing reasonable suspicion and probablecause for Zero Tolerance violations or must they remain, primarily, a lesser includedcharge of DUI?

4. When, how, and with what procedures should an officer pursue a DUI interviewfollowing a crash? How are such interviews currently being pursued? What are the cuesand what is the relative weight of each cue in reference to the driving actions related tocrash occurrence?

5. Currently available PAS units were designed to distinguish between drivers whowere at or above 0.10 percent BAC versus drivers who were below 0.10 percent. Theseunits test a measured sample of air, collected at a precise distance from the driver whilethe driver is speaking, and calculate a numeric BAC value. Units for Zero Toleranceenforcement would have different characteristics. Their function would be to detect anyalcohol without the requirement to calculate a specific value. Can Zero Tolerance units bedesigned? Similarly, can we design PAS units that will be better accepted by patrolofficers?

6. Better driver screening procedures would presumably lead to more impaireddriving arrests of nontraditional drinking driving suspects (e.g., more women and youths)and, again presumably, of suspects at BACs that are closer to the legal limit. Do thepolice really want to apprehend these “nontraditional” and “close to the limit” persons?Or would they rather spend their resources detecting and arresting the “falling downdrunk?” If we want more nontraditional and Zero Tolerance arrests, will we have to builda better rationale for these types of enforcement actions?

REFERENCES

Compton, R. P. Pilot Test of Selected DWI (Driving while Intoxicated) Detection Procedures for Use atSobriety Checkpoints. Final Report DOT HS-806-724. NHTSA, U.S. Department of Transportation,April 1985.

Federal Bureau of Investigation. Crime in the United States, 1996. Uniform Crime Reports. U.S.Department of Justice, Washington, D.C., Aug. 1997.

Fell, J. NHTSA Impaired Driving Research. Presentation at Lifesavers 16, Cleveland, Ohio, April 1998.

Ferguson, S. A., J. K. Wells, and A. K. Lund. The Role of Passive Alcohol Sensors in Detecting Alcohol-Impaired Drivers at Sobriety Checkpoints. Alcohol, Drugs, and Driving, Vol. 11, No. 1, 1995, pp. 23–30.

Jones, I. S., and A. K. Lund. Detection of Alcohol-Impaired Drivers Using a Passive Alcohol Sensor.Journal of Police Science and Administration, Vol. 14, No. 2, 1986, pp. 153–160.

Kiger, S. M., D. C. Lestina, and A. K. Lund. Passive Alcohol Sensors in Law Enforcement Screening forAlcohol-Impaired Drivers. Insurance Institute for Highway Safety, Arlington, Va., 1991.

Leaf, W. A., and D. F. Preusser. Effectiveness of Passive Alcohol Sensors. Final Report DOT HS-808-381.NHTSA, U.S. Department of Transportation, March 1996.

Preusser E-7

Lund, A. K., and I. S. Jones. Detection of Impaired Drivers with a Passive Alcohol Sensor. In Proceedingsof the 10th International Conference on Alcohol, Drugs, and Traffic Safety. Elsevier Science,Amsterdam, the Netherlands. 1987, pp. 379–382.

McKnight, A. J., E. A. Langston, J. E. Lange, and A. S. McKnight. Development of Standardized FieldSobriety Test for Lower BAC Limits (SFST-LL). National Public Services Research-Institute, Landover,Md.

Moskowitz, H. A., M. M. Burns, and S. A. Ferguson. Detecting Breath Odor from Alcohol Consumption.In Proceedings of the 14th International Conference on Alcohol, Drugs, and Traffic Safety, Vol. 2.CERMT, Annecy, France, 1997, pp. 527–531.

Preusser, D. F. Field Evaluation of Passive Alcohol Sensors. Preusser Research Group, Inc., Trumbull,Conn., 1994.

Stuster, J. W. The Detection of DWI at BACs Below 0.10. Final Report DOT HS-808-654. NHTSA, U.S. Department of Transportation, Sept. 1997.

Tharp, V., M. M. Burns, and H. A. Moskowitz. Development and Field Test of Psychophysical Tests forDWI Arrest. Report DOT HS-865-864. NHTSA, U.S. Department of Transportation, March 1981.

Ulmer, R. G., C. R. Carroll, and D. F. Preusser. Drug Evaluation and Classification (DEC) Program SiteExperience. Preusser Research Group, Inc., Trumbull, Conn., in process.

Voas, R. B. Laboratory and Field Tests of a Passive Alcohol Sensing System. Alcohol, Drugs, and Driving,Abstracts and Reviews, Vol. 4, No. 3, 1983, pp. 3–21.

Voas, R. B., E. Rhodenizer, and C. Lynn. Evaluation of the Charlottesville Checkpoint Operations: FinalReport December 30, 1983 to December 30, 1984. City of Charlottesville Police Department, NHTSAcontract DTNH22-83-C-05088, Charlottesville, Va., May 1985.

Wells, J. K., M. A. Greene, R. D. Foss, S. A. Ferguson, and A. F. Williams. Drinking Drivers Missed atSobriety Checkpoints. Journal of Studies on Alcohol, Vol. 58, No. 5, 1997, pp. 513–517.

Comments on “Identification of Alcohol Impairment on Initial Interview”

BILL TOWER

Maryland State Police Department

E-8

Police normally look at DWI detection in three phases. Phase one occurs when thevehicle is in motion and involves everything up to the point where we actually stopsomeone. Phase two entails face-to-face contact. Phase three is the prearrest screening,psychophysical testing, preliminary breath test (PBT), and so forth.

I had to think hard about the analogy Jack Stuster made about fishing and huntingin regard to officers out there on patrol, particularly because I was once an avid fishermanand hunter. Then I started looking for drunk drivers and speeders, and there really is a lotin common there. We also heard earlier about some of the unusual ways officers makecontact with drivers, for example, the bucket of chicken left on the roof of the car. Andwhen you think about it, problems with divided attention, or someone trying to do twothings at one time, are often an indication of alcohol use. We see people all the timeleaving articles of clothing on tops of cars and other things, or at night I used to seepeople driving down the road with the interior light on in the car. And believe it or not,50 percent of the time they were drunk.

I also wanted to comment on the profile of drunk drivers used by police officers. It is true that years ago, when I came on the job, we were arresting only the 0.15 and 0.20 percent BAC drivers, and these were usually white males driving down the highway inold, probably 10–15-year-old, beat-up cars with dents all over them, at midnight or later inthe evening. That is who we were arresting for drunk driving. Well, today that has reallychanged because, through our own experiences and training, we are recognizing that somany other people are drunk drivers and that we should be directing our efforts at everyone,not just one particular type of driver.

Another point I wanted to make is that traffic patrol, at least in the state police, isnot considered to be a punishment in any way whatsoever. It is our job. And more andmore local agencies are understanding that traffic patrol is a way not only to enforcetraffic laws and catch DWIs but also to make criminal arrests. One of the programs thatwe are looking at in Maryland now is criminal interdiction. We are out on the highwaystopping people for traffic violations, but we are also training our officers about what tolook for once they get up to the car. And it is resulting in huge seizures, not only of drugsbut of the cash involved with drugs, as well as guns, stolen vehicles, all kinds of differentthings. So, we are making some progress in that area, I am happy to say.

I wanted to make a comment about sobriety checkpoints too, because reallycheckpoints involve phase one and phase two DWI detection. When we set up a checkpoint,we are stopping everyone coming down the highway. The officers are trained to watchvehicles coming into the checkpoint, regardless of whether they turn away, run over oursign, or come through the checkpoint. Once we initially stop someone and the trooper orofficer walks up to the driver’s window, we are now into phase two. So those two areas arenow in play. Checkpoints—at least as far as we are concerned—are designed to be adeterrent. We do not operate checkpoints to arrest a lot of people. In fact, we would hope

Tower E- 9

that we do not arrest a lot of people, because we would prefer to deter people from drivingdrunk in the first place. And that is where we involve the media, public informationcampaigns, and that sort of thing. Typically, we are finding that in checkpoints in Marylandanyway, we make about 1 arrest for every 100 vehicles coming through the checkpoint.

Some of the areas where we have had checkpoints over the years suddenly do notqualify for checkpoints anymore, because the number of alcohol-related crashes, which isthe basis upon which we select sites, have dropped below our limit. So we know that weare doing something right when drunk-driving crashes are not occurring in those areasanymore.

Now I want to comment on some of things mentioned as indicators of alcohol use.First of all, the odor of alcohol is not a particularly good basis for assessment, unless youreally have an opportunity to smell. At sobriety checkpoints you have problems, you mayhave a breeze blowing through, or smoke from flares, things such as that. And sometimesit is very difficult to get close enough to get a good whiff of a driver’s breath. So odormay not always be helpful, although if you do smell it, then you have an excellent cue.But it is getting to that point that is not so easy. Another thing, and I am going to hit onthis a couple of times, is what appears to be either the increasing prevalence or ourimproved ability to detect drug-impaired drivers, who are obviously not going to have theodor of alcohol about them. If it is marijuana, perhaps there is an odor of marijuana onthat person. If it is PCP, there may also be an odor, or if it is an inhalant, they may havean odor of whatever chemical they are inhaling, but the odor of an alcoholic beverage, orthe lack of it, is not necessarily going to help us when it comes to drugs. So we also needto look into that area. The Drug Recognition Expert (DRE) program is going to be veryimportant if we are ever going to do an adequate job in detecting drug-impaired drivers.

As for detecting alcohol with the passive alcohol sensor, the Attorney General inMaryland has decided not to let us use the sensors at this point. We have a PBT law,which explains that officers can use a preliminary breath test as a screening device, butthat is it. We can mention the result in court and so can the defense, which they areobviously going to do if it is very low and so benefits them. The passive alcohol sensor is,I think, a very good tool if you can use it. A lot of officers, if they are going to use eitherthe PBT or the passive alcohol sensor, will try to get up there and get a sample and if it isvery low will automatically dismiss that person as not being an impaired driver. And toget back to drugs, they will not consider whether they have someone under the influenceof a very common antidepressant drug, Valium for instance, which has the identical signsand symptoms as alcohol but without the odor. So we really have to consider that. TheDREs are doing a good job in Maryland. I have looked at the results coming back fromtests of drug-impaired drivers, and marijuana is the most common drug found. Thesecond most common drug is depressants, Valium, Xanax, all kinds of barbiturates. Andthe impairment that these people are exhibiting is fairly substantial, so that, again, a lot ofofficers who stop these drivers will stick PBTs in their mouths.

All the procedures require you to look at three different clues—the inability of theeyes to smoothly track an object; nystagmus at maximum deviation, which is as far as theeye will turn to the side; and an angle of longest nystagmus prior to 45 degrees. So youlook for all these clues, and we do pretty well at picking out the drunk drivers. I heardsome discussion today about the use of nystagmus at roadside. There are some legal andconstitutional concerns about doing a formal horizontal gaze nystagmus test at roadside


without prior suspicion. I guarantee you in Maryland you would probably have a problem.So what some officers are doing out there is, when we first walk up to a car we will standjust back from the door post so the driver has to kind of look out of the corner of his eye.We do it for safety, to keep the driver off balance, but guess what we are also looking at?Nystagmus and maximum deviation. And that will be our first little clue that maybe weneed to look at this driver a little bit closer. And it really does work. This is an especiallygood technique with chronic alcohol abusers, those who are driving at 0.15 percent BACand show little evidence of impairment when it comes to walking and turns and one-legstands and things like that, because the test from what I have seen is not affected bytolerance of alcohol, that is, it is still going to be there.

The key to detection, at least in my opinion, is twofold. First of all, we have totrain the officers. We have developed these wonderful training programs, which include alot of Jack Stuster’s work with phase one DWI detection. The problem is that all ourofficers are not getting this training. You would be surprised how many officers do nothave the proper training. If they did, they would be able to detect drunk drivers muchquicker and certainly at a better rate. How we can accomplish this, I am not sure. Someofficers in some departments do not put a big emphasis on traffic patrol, especially localagencies. Big city police departments are so concerned about murders and other thingsthat they are not getting the time they need for traffic work. What we are doing in the lawenforcement community is trying to encourage them to do more with traffic control,recognizing that criminal activities are discovered in doing so.

And the second part is more contacts, as I mentioned earlier. The more vehicles thatwe can stop the better job we are going to do, the more opportunities we are going to haveto detect those drunk and drugged drivers. To get back to checkpoints just for one moremoment, in Maryland I have had some difficulties selling some of the local commanders onwhy we should use checkpoints. And here is the reasoning that I gave to them, whichseemed to work. Let us say we take 10 patrol officers and we put them out on a saturationpatrol for 4 hours. It is reasonable to assume that they would stop maybe 10 motoristsduring that 4 hours, so they have personally contacted, face to face, 100 drivers. That ispretty good, they are putting the fear of God in 100 drivers. Well, at the sobriety checkpointwe are going to take the same 10 officers and over that 3–4 hour time span we are going tostop 1,000 vehicles. So now we have personally contacted 1,000 drivers, and somewherealong the way I heard of something called the Rule of Forty. It says that for every personwho hears or sees something, over time they will tell 40 other people. So perhaps we arecontacting 40,000 people and getting the word out. Not only do you have a better chance ofarresting drunk drivers when you contact more people, but deterrence will also come intoplay. You will be able to inform more people that we are out here, we are going to conductsobriety checkpoints, our officers are trained. If you are not deterred in the first place andyou are dumb enough to drive into our sobriety checkpoint, then chances are you are goingto get caught and arrested.

Comments on “Identification of Alcohol Impairment on Initial Interview”

SUSAN A. FERGUSON


E-11

As David Preusser notes, when a police officer approaches a vehicle for the first time hedoes so with some level of suspicion regarding impaired driving. This level of suspicionwill vary depending on the circumstances leading up to the encounter. For example, ifan officer on a routine or DUI patrol has observed a driver weaving all over the road ordriving without headlights, he may already be expecting to encounter a driver who hasbeen drinking. On the other hand, if the officer is manning a sobriety checkpoint there maybe no, or very few, driving cues prior to the vehicle coming to a stop that would indicateimpairment. In this instance, the officer has to determine very quickly whether the driverhas been drinking based on the cues observable at the driver’s window. It is during thiscritical encounter at the driver’s window that the officer must decide whether to detain thedriver for further testing or send him on his way. At sobriety checkpoints this decision,even for drivers with BACs of 0.10 percent or higher, is more often to send him on his way(Ferguson et al., 1995, Wells et al., 1997). Even for dedicated DUI patrols where drivingbehavior has been observed and DUI is suspected and is the primary reason for the stop,many impaired drivers are not detained for further testing (Lund and Jones, 1987).Moreover, drivers who are missed by the officers are not a random sample of thepopulation of drinking drivers. These drivers are more likely to be female, younger,and driving without passengers (Voas and Williams, 1985; Wells et al., 1997).

In the United States, once a driver has been stopped, police officers must have areasonable suspicion of alcohol involvement before proceeding with more detailed testingto make an arrest. Research has indicated there are some fairly reliable cues, both drivingcues and cues observable at the window that, if seen, indicate impairment. According toStuster (1997) some observable driving cues, such as weaving and driving with no lightson at night, indicate with probabilities of 0.3–0.9 that the driver is impaired. The post-stop cues, such as odor of alcohol, slurred speech, slow to respond to the officer, haveeven higher probabilities (0.86–0.97) that if detected a DUI arrest will ensue.

So why, if cues such as these are so reliable, are not more impaired drivers whoare stopped by police officers detained for further testing? One possibility is that manyimpaired drivers are not exhibiting these behaviors or cues. Another possibility is thatofficers, particularly those at sobriety checkpoints where time is at a premium, are notrecognizing what could be very subtle cues. What is needed is a quick-to-administerobjective test that can determine whether the driver has been drinking, thus giving theofficer reason to monitor for further cues. The passive sensor, a technology that has beenavailable for a number of years, is one such device. Equipped with a pump that draws in asample of the driver’s exhaled breath mixed with ambient air, the sensor can provide anindication of whether the driver has been drinking, as well as approximate BAC (a higherreading indicates a higher BAC).

Research has repeatedly confirmed the effectiveness of these devices, particularly atsobriety checkpoints (Jones and Lund, 1986; Ferguson et al., 1995; Lund and Jones, 1987).


A recent study used data from the 1996 national breath survey, in which drivers were testedusing the PASIII passive sensor as well as a preliminary breath-testing device, to determinethe reliability of the device in detecting impaired drivers (Farmer et al., 1998). These resultsindicate that if a police officer were simply to obtain a sample of the driver’s breath with thePASIII when first interviewing the driver at the window, detection rates for drivers withhigh and low BACs could be improved dramatically. For example, at BACs of 0.08 percentand higher, more than 60 percent of all drivers go undetected at checkpoints (Wells et al.,1997). With the passive sensor, using a criterion of three bars as an indication that furthertesting is warranted (this device uses a series of nine colored bars to indicate approximatealcohol concentration), only 37 percent of drivers with BACs of 0.08 percent or higherwould be sent on their way. Likewise at BACs of 0.10 percent or higher, only about 25 percent of drivers would not be detained for further testing, compared with about40–50 percent without the sensors. The sensors could also be used to great effect indetecting drivers with BACs less than 0.05 percent. Using a criterion of two bars,almost a third of drivers at these lower BACs could be identified.

In spite of results such as these, passive sensors are not well liked by policeofficers for reasons outlined by Leaf and Preusser (1996). If communities are to takeseriously the challenge of enforcing 0.08 percent BACs and zero tolerance laws, then, asStuster (1997) has pointed out, they will need more than driving and post-stop cues,which typically do a poor job of singling out impaired drivers with lower BACs.

The question we need to ask ourselves is whether we are really concerned aboutthe large numbers of impaired drivers being missed. How well is it recognized amongpolice officers that so many impaired drivers who already have been stopped by police areescaping the net? Finally, do we want to expend our resources on reducing the number ofdrivers who are missed, particularly when it comes to drivers with lower BACs?

REFERENCES

Farmer, C. M., J. K. Wells, R. B. Voas, and S. A. Ferguson. Field Evaluation of the PAS III PassiveAlcohol Sensor. Journal of Crash Prevention and Injury Control, 1998.


Jones, I. S., and A. K. Lund. Detection of Alcohol-Impaired Drivers Using a Passive Alcohol Sensor.Journal of Police Science and Administration, Vol. 14, No. 2, 1986, pp. 153–160.

Leaf, W. A., and D. F. Preusser. Effectiveness of Passive Alcohol Sensors. Final Report DOT HS-808-381.NHTSA, U.S. Department of Transportation, March 1996.

Lund, A. K., and I. S. Jones. Detection of Impaired Drivers with a Passive Alcohol Sensor. In Proceedingsof the 10th International Conference on Alcohol, Drugs, and Traffic Safety, Elsevier Science,Amsterdam, the Netherlands, 1987, pp. 379–382.

Stuster, J. W. The Detection of DWI at BACs Below 0.10. Final Report DOT HS-808-654. NHTSA, U.S.Department of Transportation, Sept. 1997.

Voas, R. B., and A. F. Williams. Age Differences of Arrested and Crash-Involved Drinking Drivers.Journal of Studies on Alcohol, Vol. 47, No. 3, 1985, pp. 244–248.

Wells, J. K., M. A. Greene, R. D. Foss, S.A. Ferguson, and A.F. Williams. Drinking Drivers Missed atSobriety Checkpoints. Journal of Studies on Alcohol, Vol. 58, No. 5, 1997, pp. 513–517.

Identification of Impairment Outside theVehicle: Field Sobriety Tests

MARCELLINE M. BURNS

Southern California Research Institute

ABSTRACT

The Standardized Field Sobriety Test (SFST) battery includes three tests: horizontal gazenystagmus (HGN), walk and turn, and one-leg stand. Following the 1970s research thatidentified these best tests, a training curriculum was developed by the National HighwayTraffic Safety Administration (NHTSA). Officers in 50 states now have been trained underNHTSA guidelines with the SFSTs, which have been revalidated for 0.08 percent bloodalcohol concentration. In data obtained during recent field studies with officers trained andexperienced with the test battery, arrest decisions were found to be more than 90 percentcorrect. Defense attorneys vigorously oppose the admissibility of HGN testimony in DUItrials, and courts in 38 states have considered admissibility and related issues. The rulings aremixed but largely favorable. The SFSTs remain an important tool for alcohol enforcement.

INTRODUCTION

In 1939 the State of Indiana enacted the first statute in North America defining a DUIoffense in terms of blood alcohol concentration (BAC), specifically 0.15 percent. Mainepassed a similar law shortly thereafter. The precedent for this kind of legislation had beenset by Norway in 1936. Although it now is difficult to determine whether in the UnitedStates those first statutory limits were enforced effectively, it is certain that at least twofactors would have had an impact on the efforts that police officers brought to the task.Both the state-of-knowledge at that point in time about alcohol’s effects on driving andthe particular alcohol level set by the statute would have affected enforcement.

Intensive alcohol-and-driving research had not yet begun when Indiana passedthe 0.15 percent law, which means there was substantially less information about theconsequences of combining alcohol and drugs than is available in 1998. Assuming that thelevel sanctioned by the statute represented a consensus of legislators and citizens, if notscientists, it is likely that the lack of scientific data about the effects of lower BACscontributed to the point of view. The 0.15 percent level also set the stage for enforcementproblems that it now appears were not fully recognized for several decades.

A law allowing driving with BACs up to 0.15 percent indisputably was a drunk-driving law and, as such, it limited traffic officers’ alcohol enforcement responsibilities tothe detection and arrest of drunk drivers. Although the state of being “drunk” has noprecise scientific meaning, it typically is equated with obvious intoxication, and all but thetolerant drinker do indeed typically appear “drunk” at 0.15 percent. Retrospectively, itseems likely that enforcement practices were shaped for several decades by the belief thattraffic officers would be able to easily recognize the obvious intoxication of drunk drivers.

F-1

F-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

Over time, however, research findings from the laboratory and from the fieldserved to change both the duties and the needs of traffic officers. As accumulating dataconvincingly demonstrated impairment of driving skills at lower BACs, the definitionof the problem shifted from drunk driving to alcohol-impaired driving. Legislatorschanged the laws, and traffic officers’ responsibilities expanded beyond drunk drivers toencompass the detection and arrest of alcohol-impaired drivers. As a result of the shift, adecrease in the mean BAC of arrested drivers certainly could have been expected, but thatexpectation was thwarted. The mean BAC of arrested drivers remained at 0.17 percenteven after 0.10 percent statutes had been enacted across the nation.

Officer proficiency played a role in that statistic. At the time, neither officers northeir departments uniformly viewed alcohol-impaired driving as a high priority area ofenforcement, and resources were preferentially allocated to other areas that were seen asreal crimes. The 0.17 percent mean BAC statistic itself is evidence that officers lackedthe skills, and perhaps the will, to detect and arrest drivers with lower BACs. Althoughthe statutes had been changed, the payoff for highway safety awaited effectiveenforcement, which in turn awaited changes in police department policies and trainingacademy curricula. Beginning in the mid-1970s NHTSA launched field sobriety testresearch, which eventually would effect the needed changes in curricula.

SOBRIETY TEST RESEARCH

Early Laboratory Studies

Because early traffic safety literature is notably lacking in reports of standardized testing,the first large studies of sobriety tests, which were conducted by Finnish investigatorsduring the 1970s, are of particular interest (Penttila, Tenhu, and Kataja, 1971, 1974). Theirresearch reports serve as a kind of overview of the state of sobriety testing for that time, notonly for Finland but for the United States as well. Notice in Tables 1 and 2 the similaritybetween the list of tests examined in Finland and the list of tests observed at roadside in theUnited States (Burns and Moskowitz, 1977). Although language differences producesomewhat different test names, the test categories are largely the same.

Clinical examinations of suspected drunken drivers in Finland were conducted bypolice surgeons, and the research reports indicate that testing was carried out with somedegree of consistency (Penttila et al., 1971, 1974). In contrast, the testing conducted bypolice officers in the United States differed between agencies, between officers, and evenbetween suspects (Burns and Moskowitz, 1977). Although officers’ roadside test repertoirestypically included some combination of balance, walking, finger dexterity, alphabetrecitation, and/or miscellaneous maneuvers, there was little consistency either as to whichtests were chosen or how the chosen ones were administered.

In general, the concept of test standardization had not yet entered either thecurricula of law enforcement academies or the practices of officers on patrol. Oneexception to variability in testing that did occur was in connection with systems forvideotaping the examination of suspects. To aid in the prosecution of drivers chargedwith DWI (or DUI),1 agencies equipped a stationhouse room with cameras and

1 Driving under the influence (DUI) and driving which intoxicated (DWI) are used interchangeably to mean driving witha BAC that violates the jurisdiction’s per se or presumptive statute.

Burns F-3

equipment for the purpose of making evidential records of breath testing and sobrietytesting. Officers in those locations generally used the same tests and procedures for allsuspects.

Southern California Research Studies

As part of a broader research program by NHTSA to address alcohol issues, the SouthernCalifornia Research Institute (SCRI) conducted two large experiments for the specificpurpose of identifying a best set of sobriety tests (Burns and Moskowitz, 1977; Tharp,Burns, and Moskowitz, 1981). The procedures of the two SCRI studies were similar; asummary follows.

As an initial effort, a list of candidate tests of sobriety was compiled by searchingfor alcohol-sensitive psychomotor tests, by broadly reviewing the test and measurement

TABLE 1 Clinical Tests for ExaminingSuspected Drunken Drivers in Finland

TABLE 2 Sobriety Tests Observed atRoadside in the United States in 1975

SOURCE: Penttila, Tenhu, and Kataja (1974).


literature as well the alcohol and traffic literatures, and by riding with and observingroadside activities of officers on traffic patrol in several states. The kind of test that couldbe considered was limited by both the unique requirements and the constraints imposedby roadside environments. The fundamental criterion of alcohol sensitivity was itselfunusually difficult, given that officers rarely have information about the soberperformance of the drivers they stop. Also, since a driver cannot be detained on suspicionfor an unreasonable time period, only relatively brief tests were eligible. Tests requiringspecial conditions were ruled out, because the terrain and weather conditions of roadsidestops usually are not subject to control. Tests that required bulky or complex apparatuswere not considered, because they would not have been operationally acceptable toofficers, nor would it have been feasible to recommend tests that required policedepartments to purchase costly equipment. Also, the performance demands of an alcohol-impairment test, by definition, cannot exceed the performance capabilities of an averageperson when he or she is not alcohol-impaired. Finally, to be acceptable to police officers,the attentional demands of administering the tests had to be simple enough that the safetyof officer, suspects, and bystanders would not be compromised in any way.

An initial, relatively short list of candidate tests was compiled. Many of thosewere eliminated during pilot testing with the result that only six tests, together with fouralternates, were identified for examination in a full-scale experiment. The six tests werehorizontal gaze nystagmus (HGN), walk and turn (WAT), one-leg stand (OLS), finger-to-nose, finger count, and a paper-and-pencil tracing test. The alternate tests were Romberg(body sway), subtraction, counting backward, and letter cancellation.

More than 450 men and women, including light, moderate, and heavy drinkers,participated in the two experiments. In order for the sample to be representative of driverscommonly encountered by police officers, the only criteria applied to subjects were: ages21 years or older, licensed drivers, and self-reported alcohol users. Police officers, 10 foreach experiment, were recruited through local law enforcement agencies to serve asexaminers. They were trained to administer and score the selected tests in a standardizedmanner during a single, four-hour session scheduled during the week prior to the officers’participation.

Subjects consumed alcohol beverages over a 90-minute period and reachedmeasured BACs ranging from 0.00 percent to 0.15 percent. In the first experiment,officers administered six tests to each subject and recorded an arrest/don’t arrest decisionbased on their observations of test performance. Analysis of the data identified HGN,WAT, and OLS as a best set of tests. A second experiment with essentially identicalprocedures was conducted for the purpose of further refining standardized administrationand scoring of the recommended three-test battery. It is interesting to note that althoughthe BAC limit in California was 0.10 percent at the time of the studies, analysis of theofficers’ decisions showed that the BAC at which they predominantly made arrestdecisions was 0.08 percent.

Test selection for the first experiment took into account that physical impairments,which do not interfere with an individual’s driving ability, may nonetheless preclude theexamination of sobriety with tests of balance and walking. With a single exception,however, tests that did not require balance and walking did not meet criteria for inclusionin a battery. The exception was HGN, which requires only that the suspect move his orher eyes as directed. It is noteworthy too that although the officers lacked experience and

Burns F-5

confidence with HGN, it emerged even during the early development of the SFST batteryas the best index of alcohol.

Following a report of the second SCRI study, NHTSA personnel developed ascoring method, training guidelines, student and instructor manuals, and other trainingmaterials. The concept of stringent adherence to standardized testing was incorporatedinto all of these materials from the beginning and continues to be emphasized throughouttraining. Officers are taught exact procedures for verbal test instructions, testdemonstrations, and scoring. In training programs under NHTSA guidelines, trafficofficers in all 50 states have been trained to use the three tests, which now are known asSFSTs. The tests are also part of the Drug Recognition program’s 12-step protocol forexamining drivers suspected of drug impairments.

SFSTs for Low BACs

At the time of the field sobriety test research by SCRI, drivers were subject to arrest at0.10 percent BACs. Thus, questions about valid tests for roadside use were raised anewby the enactment of laws specifying 0.08 percent for drivers in general, 0.04 percent forcommercial drivers, and lower limits for youthful drivers. Again, NHTSA contractresearch addressed the issue. McKnight et al. (1995) conducted a laboratory study todevelop sobriety tests for lower BAC limits and concluded the current SFSTs are as validfor 0.08 percent BAC as for 0.10 percent BAC. They also reported that HGN withmodified scoring is valid for lower BACs.

Field Studies

Laboratory studies identified, standardized, and validated the battery of SFSTs, whichhave been widely accepted by law enforcement. It has been assumed that officersgenerally have had positive experience with the tests and believe their use leads to morecorrect decisions than would be possible otherwise. Until data were obtained from actualroadside decisions, however, the accuracy associated with the SFSTs was unknown.

The second SCRI laboratory study included a small field study, which wasconducted in collaboration with deputies from the Los Angeles County Sheriff’s Office.Due to the limited scope and duration of that study, it supported only very generalconclusions, and questions about use of the SFSTs under real-world conditions remainedlargely unanswered.

Anderson et al. (1983) examined the feasibility of field use of the battery with astudy in four jurisdictions. Based on officers’ administrations of the test battery to DWIsuspects, the investigators concluded that field data confirmed the findings from the SCRIlaboratory studies. They reported that the battery was about as effective as preliminarybreath testers (PBTs) in reducing false positive arrests, and they found gaze nystagmus tobe the most powerful of the three tests. The findings from the study were weakenedbecause officers’ decisions could have been influenced by knowledge of BACs. In amajority of the cases, the alcohol levels measured with a PBT were known to the officersprior to arrests.

Findings reported from a field study of the SFSTs by the Vermont AlcoholResearch Center are mixed but largely negative (Perrine et al., 1993). Study participants


were 361 drivers encountered at roadside and 119 dosed subjects. Four police officers,described as SFST experts, administered the tests to 185 subjects. Six Ohio policeofficers, described as having had NHTSA-specified training in SFST procedures,administered the tests to 93 subjects. The Ohio officers were further described as being orhaving been active in DUI enforcement. Research assistants tested 202 subjects.

The investigators concluded that well-trained, experienced, and highly motivatedtest administrators can use HGN with some accuracy at BACs of 0.05 percent, 0.08 percent,and 0.10 percent. They stated further that their data demonstrate limited validity for thebattery, especially WAT and OLS, and they suggested that either the entire battery, or onlyWAT and OLS, be abandoned in favor of “chemical-based” enforcement.

If indeed the SFSTs cannot be shown to serve as valid and reliable indices ofalcohol, then the current level of confidence vested in them by law enforcement and bysome courts amounts to very serious error. In the interest of traffic safety, the findings ofPerrine et al. (1993), which differ from those reported by other investigators, must begiven careful attention.

The relevance of the VARC data specifically to use of the SFSTs by policeofficers is weakened somewhat by the inclusion of data for 202 (42 percent)administrations of the test battery by research assistants. Although the report describes“mandatory training” of the research assistants by SFST “experts,” it is unclear whethertheir training was equivalent to that of police officers, that is, under NHTSA guidelinesby certified SFST instructors. Also, following training, the proficiency of police officerstypically shows a steep learning curve as they hone their newly acquired skills anddevelop confidence in their decisions based on the tests. Without information about theresearch assistants’ total experience with test administration and the criteria applied toassess the adequacy of their testing skills, it is unknown whether the reported findingsreflect deficiencies of the tests or deficiencies of the examiners.

Additional field data now are available. The question, “How accurate are thedecisions made by experienced, skilled officers under roadside conditions when theyrely on the SFSTs?” was addressed by two recent studies in which SFSTs wereadministered at roadside by traffic officers trained under NHTSA guidelines. As can be seen in the following, the data led to different conclusions than those of Perrine et al. (1993).

A study was conducted in collaboration with seven law enforcement agencies inthe State of Colorado during the period February to July 1995 (Anderson and Burns,1997). Trained and experienced officers were given refresher training by a certifiedinstructor. The officers relinquished the PBTs commonly in their possession and agreedto use no roadside test other than the SFSTs throughout the study period. At theconclusion of data collection, they signed written certifications that they had compliedwith these and all other study conditions.

Observers were an important component of the study. A group of individualsassociated with law enforcement either as sworn officers or civilian employees weretrained in study procedures. They were assigned by random procedures to ride with thestudy officers and were present for 41 percent of the SFST administrations. Theyobserved and reported whether officers adhered to study and testing procedures. Theyalso requested and obtained breath specimens with preliminary breath testing instrumentsfrom drivers who performed the SFSTs but were not arrested. Drivers refused to provide

Burns F-7

a breath specimen for 10 percent of the requests. These data permitted calculation ofBAC estimates for released drivers.

The records, including breath or blood test results, were tracked, retrieved, andanalyzed for all roadside stops that involved administrations of the SFSTs (N = 305).BACs supported 93 percent of the officers’ arrest decisions. Under Colorado’s two-tieredstatute, 30 arrested drivers whose measured BACs were in the range 0.05 percent to0.095 percent were charged with Driving While Alcohol Impaired (DWAI). The BACs of133 arrested drivers were 0.10 percent or higher, and they were charged with DrivingUnder the Influence (DUI). Since drug evaluations were not conducted, there was nodetermination of impairment for the drivers whose BACs did not support an arrest.

Approximately one-third of the drivers who were released at roadside should havebeen arrested. The records for each of those cases were examined separately and as mightbe expected, it was found that lower BACs generated many of the release errors. Roughlytwo-thirds of the incorrectly released drivers were below 0.10 percent and, therefore,were subject to arrest under the lesser charge. A number of those were only slightlyabove 0.05 percent.

A second study was carried out in Florida in cooperation with the Pinellas CountySheriff’s Office (Burns and Dioquino, 1997). The study procedures replicated those of theColorado study. Again, the officers’ decision were found to be largely correct. Breath testresults supported 95 percent of the deputies’ decisions to arrest and 82 percent of thedecisions to release.

A third recent study conducted by Anacapa Science in collaboration with the SanDiego Police Department also examined the validity of the SFSTs (Stuster and Burns, 1997).In this study, officers estimated whether a driver’s BAC was above or below 0.08 percent or0.04 percent, and their estimates were more than 90 percent correct.

HORIZONTAL GAZE NYSTAGMUS

HGN has been found to be the most sensitive and reliable test in the three-test battery. Ithas also been the most controversial. Although the entire SFST battery has been targetedrepeatedly by court challenges in the State of Florida, challenges in other jurisdictionstypically have been directed solely to HGN.

Nystagmus can be simply defined as a jerking motion of the eyeball (Aschan et al., 1956; Aschan, 1958; Richman and Jakobowski, 1994). The several types includethose with a vestibular system origin (e.g., positional alcohol nystagmus, caloricnystagmus, post-rotational nystagmus) and those with a neural origin. Gaze nystagmus,or HGN, is the latter type and also is the only nystagmus recommended for use as aroadside test. It is evoked by causing the eyes to move in the lateral plane.

Three signs or characteristics of HGN are associated with the presence of alcohol,other CNS depressants, inhalants, or phencyclidine. They are (1) a lack of smoothpursuit, (2) a distinct jerking at maximum deviation of the eyes, and (3) an onset ofjerking prior to a 45° angle of gaze (AOG). The procedures for observing a suspect’s eyesfor HGN include the following:

• Ensure that (1) the suspect can see the stimulus and (2) you can see the suspect’seyes;


• Observe each eye separately for the three HGN signs;• Hold a stimulus object (pen, penlight, or finger) slightly elevated above the line of

sight and approximately 15 inches in front of the suspect’s face; and• Score one point for one sign observed in one eye (three points per eye; possible

six point total).

A total score of four points is evidence that the suspect’s BAC is ≥ 0.10 percent.The HGN examination to determine whether a suspect’s eyes track smoothly

begins with a stimulus held at center vision directly in front of the subject’s face. Thestimulus then is moved in a straight pathway to the extreme left, back to center, then tothe extreme right and back to center. The movement can be repeated as necessary toconfirm an observation. If an HGN-producing substance is present, the suspect’s eyeswill exhibit a jerky movement as they pursue the stimulus.

For examination of the second HGN sign, a stimulus is moved to the side todeviate the lateral position of the suspect’s eyes as far as possible. Because the strain ofthe position may produce a slight tremor of the eyeball, the stimulus must be held at thatposition for approximately four seconds. The sign, jerking at maximum deviation, isscored only if it is distinct and persists for the entire four seconds.

The third sign is the AOG at the onset of the jerking movement. This importantmeasure reflects the relationship of AOG to BAC. As the BAC rises, the jerkingcorrespondingly occurs earlier, that is at a lesser angle of deviation. The requiredaccuracy of observation dictates a slightly slower movement of the stimulus from thecenter position toward the periphery. A point is scored whenever the jerking movementoccurs prior to a 45° AOG.

The medical literature about nystagmus as a sign of alcohol dates back at leastseveral decades. A study by Lehti (1976) reported that the angle of gaze and BACcorrelate highly significantly. Following the introduction of the SFSTs, investigatorsbegan to examine law enforcement’s use of HGN. Goding and Dobie (1986), reported astudy of 46 emergency room patients and 159 DWI suspects. They found no cases ofBAC overestimation by gaze nystagmus and concluded that it is an effective tool forestimating BAC. They cautioned, however, that observations may be contaminated by thepresence of other drugs.

Good and Augsburger (1986) examined 2,429 records of SFST administration byOhio State Highway Patrol officers. All of the drivers were arrested, and the BACs of2,223 (91.5 percent) were found to be 0.10 percent or higher. These investigatorsconcluded that HGN is the most accurate roadside test of sobriety, and that it will enablepolice officers to make more informed arrest decisions.

Richman and Jakobowski (1994) examined the use of HGN by newly trainedpolice officers. Thirty recruits were trained during a two-day course by NHTSA-certifiedinstructors, experienced Massachusetts police officers, and state troopers. Each traineethen made HGN observations with seven subjects (N = 210 examinations) whose BACsranged from 0.00 percent to 0.11 percent. From an analysis of pass/fail decisions, theinvestigators reported 87.78 percent sensitivity (correct failures) and 86.67 percentspecificity (correct passes). In still another study, in which cognitive tests were evaluatedin comparison to the SFST battery, nystagmus emerged as the best alcohol predictorfollowed by code substitution (Kennedy et al., 1994).

Burns F-9

The criminal defense community regularly publishes legal documents arguingagainst HGN (e.g., Cowan and Jaffee; Rouleau; Busloff, 1995). For the most part, thesepublications consist of legal arguments and interpretations of extant research. They havenot reported additional research or new data.

A number of different legal issues and arguments, which may not be entirelyunderstood by non-legal observers, are commonly advanced in court proceedings. As ofJune 1998, the admissibility of HGN evidence had been directly considered by courts in38 states, and Table 3 briefly summarizes the decisions produced by those hearings andtrials (Gould, 1998). Up-to-date summaries of HGN case law can be obtained from theNational Traffic Law Center.2

CONCLUSIONS

Alcohol enforcement by traffic officers would be relatively easy if they needed to stoponly those who commit obvious driving errors and to arrest only those who are obviouslyintoxicated. Since it may be the case, however, that alcohol impairment generates neithergross driving errors nor stereotypical drunk behavior, in reality an officer’s task often isvery difficult.

The SFSTs, in particular HGN, are important tools for an officer, because he mustmake an arrest or release decision within a few minutes of having stopped a vehicle. Thedecision cannot be delayed, and it will be made based on the sum of information availablefrom observation, interrogation, and testing. Recent field studies indicate that arrestdecisions made by officers trained and experienced with the SFSTs are more than 90 percentcorrect. It is instructive to further consider roadside activities and the decision process fromthe point of view of an officer.

TABLE 3 Horizontal Gaze Nystagmus CourtRulings

2 National Traffic Law Center, 99 Canal Center Plaza, Alexandria, VA 22314, (703) 549-4253.

SOURCE: National Traffic Law Center


• He stops a vehicle.• His initial observations lead him to suspect alcohol impairment.• He instructs the driver to get out of his vehicle, and he administers the SFSTs.• If he decides to arrest, he transports the driver to a booking facility and obtains a

BAC.• If the BAC is measured by breath specimen, as it most often is, he knows

immediately whether his roadside decision was correct or incorrect.

Given this kind of regular feedback, the widespread acceptance of the SFSTsconstitutes compelling evidence of their validity. It is unlikely in the extreme either thatan officer would continue to use, or that a law enforcement agency would permitcontinued use of, tests that repeatedly produce error. Handcuffing, transporting, andbreath-testing a driver takes an officer off active patrol for a period of time. Wheneverthat driver then is found not to be alcohol-impaired, officer time has been wasted and acitizen has been inappropriately harassed.

The DUI defense community’s vigorous efforts to keep testimony about theSFSTs out of court proceedings also speak to the issue. HGN continues to be targetedfrequently as an admissibility issue. If the tests produce error, then testimony about themsurely would be detrimental to prosecution. If that were occurring, it is unlikely thatdefense attorneys would exert legal time and effort to ban such testimony.

Perhaps it is most telling that opponents of the SFSTs have not offered alternativetests. Officers must make timely decisions about alcohol impairment based on whateverinformation is available to them at roadside. At the present time, no other set of tests hasbeen shown to lead to better decisions. Unless and until laws, policies, and practicesallow DUI charges to be filed based solely on chemical tests, continued use of the SFSTsappears to be the only and the best option.

REFERENCES

Anderson, T. E. Development of Effective Behavioral Test Procedures for Alcohol-Impaired DriverIdentification. Research Notes, NHTSA, U.S. Department of Transportation, 1983.

Anderson, E., and M. Burns. Standardized Field Sobriety Tests: A Field Study. In Proceedings of the 14thInternational Conference on Alcohol, Drugs, and Traffic Safety, Vol. 2, 1997, pp. 635–639. CERMT,Annecy, France.

Anderson, E., and M. Burns. Standardized Field Sobriety Tests: A Field Study. The DRE, Vol. 9, No. 3,1997, pp. 6–8.

Anderson, T., R. Schweitz, and M. Snyder. Field Evaluation of a Behavioral Test Battery for DWI.Technical Note DOT HS-806-475. NHTSA, U.S. Department of Transportation, 1983.

Aschan, G. Different Types of Alcohol Nystagmus. Acta Otolaryngology (Suppl), Vol. 140, 1958, pp. 69–78.

Aschan, G., M. Bergstedt, L. Goldberg, and Laurell, L. Positional Nystagmus in Man during and afterAlcohol Intoxication. Quarterly Journal Studies Alcohol, Vol. 17, 1956, pp. 381–405.

Belton, H. Lateral Nystagmus: A Specific Diagnostic Sign of Ethyl Alcohol Intoxication. New ZealandMedical Journal, Vol. 100, 1987, pp. 534–535.

Bender, M., and F. O’Brien. The Influence of Barbiturates on Various Forms of Nystagmus. AmericanJournal of Ophthalmology, Vol. 29, 1946, pp. 1541–1552.

Burns F-11

Burns, M. Sobriety Tests for the Presence of Drugs. Alcohol, Drugs, and Driving, Vol. 3, No. 1, 1987,pp. 25–29.

Burns, M. Horizontal Gaze Nystagmus: The Controversy and the Issues. The DRE, Vol. 3, No. 3, 1991.

Burns, M., and T. Dioquino. Florida Validation Study of the Standardized Field Sobriety Test (SFST) Battery.Project No. AL-97-05-14-01, State Safety Office, Department of Transportation, State of Florida, 1997.

Burns, M., and P. Gould. Police Evaluation of Alcohol and Drug Impairment: Methods, Issues, and LegalChallenges. In Proceedings of the 14th International Conference on Alcohol, Drugs, and TrafficSafety, Vol. 2. CERMT, Annecy, France, 1997, pp. 629–634.

Burns, M. M., and H. A. Moskowitz. Psychophysical Tests for DWI Arrests. Report DOT HS-802-424.NHTSA, U.S. Department of Transportation, 1977.

Busloff, S. E. Can Your Eyes Be Used Against You? The Use of the Horizontal Gaze Nystagmus Test inthe Courtroom. The Journal of Criminal Law and Criminology, Vol. 24, No. 1, 1995, pp. 203–238.

Compton, R. P. Pilot Test of Selected DWI Detection Procedures for Use at Sobriety Checkpoints. ReportDOT HS-806-724. NHTSA, U.S. Department of Transportation, 1985.

Cowan, J. D., and S. G. Jaffee. Proof and Disproof of Alcohol-Induced Driving Impairment throughEvidence of Observable Intoxication and Coordination Testing. 9 Am Jur Proof of Facts 3d, 459–492.

Dell’osso, D. P., R. B. Daroff, and B. F. Troost. Nystagmus and Saccadic Intrusions and Oscillations. InDuane’s Clinical Ophthalmology, Vol. II, (W. Tasman and E. A. Jaeger, eds.). Lippincott, Pa., 1989,pp. 1–30.

Goding, G. S., and R. A. Dobie. Gaze Nystagmus and Blood Alcohol. Laryngoscope, Vol. 96, No. 7, 1986,pp. 714–717.

Good, G. W., and A. R. Augsburger. Use of Horizontal Gaze Nystagmus as Part of Roadside SobrietyTesting. American Journal of Optometry and Physiological Optics, Vol. 63, No. 6, 1986, pp. 467–471.

Gould, P. What’s New in DRE/HGN Court Cases. Workshop Materials. The Fourth IACP DRE, Drugs,Alcohol, and Impaired Driving Training Conference, Portland, Ore., 1998.

Halperin, E., and R. L. Yolton. Is the Driver Drunk? Oculomotor Sobriety Testing. Journal of the AmericanOptometric Association, Vol. 57, 1987, pp. 654–657.

Kennedy, R. S., J. J. Turnage, and G.G. Rugotzke. Indexing Cognitive Tests to Alcohol Dosage andComparison to Standardized Field Sobriety Tests. Journal of Studies on Alcohol, 1994, pp. 615–628.

Kosnoski, E. M., R. L. Yolton, K. Citek, C. E. Hayes, and R. B. Evans. The Drug Evaluation ClassificationProgram: Using Ocular and Other Signs to Detect Drug Intoxication. Journal of the AmericanOptometric Association, Vol. 69, No. 4, 1998, pp. 211–227.

Lehti, H. M. J. The Effect of Blood Alcohol Concentration on the Onset of Gaze Nystagmus. Blutalkohol,Vol. 13, 1976, pp. 411–414.

McKnight, A. J., E. A. Langston, J. E. Lange, and A. S. McKnight. Development of Standardized FieldSobriety Test for Lower BAC Limits (SFST-LL). Contract No. DTNH22-92-97000. NHTSA, U.S. Department of Transportation, 1995.

McKnight, A. J., and E. A. Langston. The Use of Video in Training for Standardized Field Sobriety Tests(SFST). Contract No. DTNH22-92-R-05109. NHTSA, U.S. Department of Transportation, 1993.

Penttila, A., M. Tenhu, and M. Kataja. Clinical Examination for Intoxication in Cases of SuspectedDrunken Driving. Statistical and Research Bureau of TALJA, Iso Roobertinkatu 20, Helsinki 13,Finland, 1971.

Penttila, A., M. Tenhu, and M. Kataja. Examination of Alcohol Intoxication in Cases of Suspected DrunkenDriving II: A Mathematical Analysis of the Relationship Between the Results of Clinical Examinationand Blood Alcohol. LIIKENNETURVA, Iso Roobertinkatu 20, 00120 Helsinki 12, Finland, 1974.


Perrine, M. W., R. D. Foss, A. R. Meyers, R. B. Voas, and C. Velez. Field Sobriety Tests: Reliability andValidity. In Proceedings of the 12th International Conference on Alcohol, Drugs, and Traffic Safety,1993, pp. 1133–1138. Verlag TUV Rheinland, Cologne.

Rashbass, C. Barbiturate Nystagmus and the Mechanisms of Visual Fixation. Nature, Vol. 183, 1959, pp. 897–898.

Richman, J. E., and J. Jakobowski. The Competency and Accuracy of Police Academy Recruits in the Useof the Horizontal Gaze Nystagmus Test for Detecting Alcohol Impairment. New England Journal ofOptometry, Vol. 47, No. 1, 1994, pp. 5–8.

Rouleau, M. A. Unreliability of the Horizontal Gaze Nystagmus Test. 4 Am Jur Proof of Facts 3d, pp. 439–495.

Simpson-Crawford, T., and S. Slater. Eye Signs in Suspected Drinking Drivers: Clinical Examination inRelation to Blood Alcohol. New Zealand Medical Journal, Vol. 74, 1971, pp. 92–96.

Stuster, J., and M. Burns. Validation of the Standardized Field Sobriety Test Battery at BACs below 0.10Percent. Contract No. DTNH22-95-C-05192. NHTSA, U.S. Department of Transportation, 1997.

Tharp, V. Gaze Nystagmus as a Roadside Sobriety Test. Alcohol, Drugs, and Driving, Abstracts andReviews, Vol. 2, 1981, No. 2, pp. 5–8.

Tharp, V., M. M. Burns, and H. A. Moskowitz. Development and Field Test of Psychological Tests forDWI Arrests. Report DOT HS-805-864. NHTSA, U.S. Department of Transportation, 1981.

Umeda, Y., and E. Sakata. Alcohol and the Oculomotor System. Annals of Otology, Rhinology andLaryngology, Vol. 87, No. 3, 1978, pp. 392–397.

Comments on “Identification of Impairment Outside the Vehicle”

ROBERT B. VOAS

Pacific Institute for Research and Evaluation

F-13

Dr. Marcy Burns has provided an excellent summary of the development, application,and effectiveness of field sobriety tests. For the first three quarters of this century, policedepended on observations that could be made by most adults. Police arrested individualswho were obviously intoxicated based on traditional signs such as staggering, slurredspeech, incoherence, and loss of manual dexterity. When the breath test first came on thescene, the blood alcohol concentration (BAC) limit was 0.15, a level at which mostdrinkers will demonstrate some of these traditional signs of impairment. Once the BAClimits were reduced to 0.10 or 0.08, the police were required to adopt more refinedobservational techniques. The development of field sobriety tests has served well for thispurpose. As Dr. Burns has noted, field studies of the accuracy of field sobriety tests haveshown them to be valid when used by properly trained officers.

IMPAIRED DRIVERS MISSED BEFORE STANDARD FIELD SOBRIETY TESTS (SFSTs)

Not revealed by these studies, however, is the dark figure of the high BAC drivers whoare not detected and, therefore, not subjected to sobriety testing. To be a candidate forsobriety testing, the suspect must have shown some unusual driving behaviors thatcaused the police to stop him or her. During the interview at the driver’s window, thesuspect must also appear to have consumed a significant amount of alcohol. Drivers whomight be over the BAC limit but who do not meet both of these criteria cannot berequired to leave the vehicle and submit to a sobriety test. The extent to which these twoscreening requirements result in missing high BAC drivers whose impairment is notreadily apparent is demonstrated from research on the use of passive sensors at sobrietycheckpoints (Lund and Jones, 1987; Ferguson et al., 1995) and in regular patrols(Taubenslag and Taubenslag, 1975). In general, half of the drivers in these studies withBACs over 0.10 were not detected. Thus, there is evidence that, at the current per selimits of 0.08 and 0.10, approximately one-half of the impaired drivers will be missedby an enforcement system that is based almost entirely on observations of driving,personal appearance, and motor behavior.

An example of how the current system is operating is provided by the state ofCalifornia, which has had an 0.08 law since 1990 and boasts a reputation for strongenforcement. Despite this, the average BAC of drivers arrested for driving under theinfluence (DUI) by its well-equipped, highly trained state patrol in 1996 was 0.16, aboutthe same as other large states, which vary from 0.13 to 0.18 (MADD, Rating the States,1996). While 68.5 percent of the motorists in weekend nighttime roadside surveys whohave BACs over 0.08 fall in the interval between 0.08 and 0.12, only 30.6 percent ofthose arrested fall into that interval. Thus, even the best enforcement organizations have astrong bias toward the high end of the BAC scale. Since crash risk is a function of BAC,


this bias can be defended. However, it clearly shows that our current system isconsiderably less efficient at the marginal BAC levels.

As noted, DUI enforcement techniques had to be refined when BAC limits werelowered from 0.15 to 0.10 or 0.08. Now, we are further challenged with the establishmentof 0.04 limits for commercial drivers and 0.02 for underage drivers. At these lower levels,little or no behavioral impairment is to be expected. McKnight et al. (under review) haveshown, however, as Dr. Burns has reported, that horizontal gaze nystagmus can stillprovide a useful indicator at a BAC of 0.04, providing the scoring is adjusted for the lowerBAC limit. Because individuals at a BAC of 0.04 will show so few signs of impairment,McKnight demonstrated that the horizontal gaze nystagmus test could be administeredwhile the driver is seated in the vehicle. Seated administration is important because, with aBAC of 0.04, without this test the officer might not have enough evidence to require thatthe suspect exit the vehicle for a field sobriety test. Clearly, at these low levels, behavioralobservations are likely to be far less useful than at the 0.10 level. Effective enforcement isgoing to be dependent on a more aggressive use of alcohol sensors.

With the passage by all 50 states of the zero tolerance law, apprehending driversyounger than age 21 with any measurable amount of alcohol has become a nationalissue. When the National Highway Traffic Safety Administration (NHTSA, 1991)conducted a national survey of police officers as part of its report to Congress on BAClimits, considerable concern was expressed about enforcing low BAC limits of 0.04 to0.06. Most police, however, felt that enforcing a zero level would be considerably easiersince any evidence of drinking would be the basis for a citation. The extent to which thezero tolerance law will be enforced remains to be determined. Preusser et al. (1992)have identified several problems in enforcing drinking and driving laws with underagedrivers beyond the detection of a low BAC that may reduce the number of citationsissued under these laws. One likely complication is that while the law prohibits drinkingper se, officers will continue to rely on finding impairment. They may continue to relyon field sobriety tests even though no measurable impairment can be expected in youthswho have had one or two drinks. Of course, where the officer has evidence of drinkingand plans to make a charge under the zero tolerance law, conducting a field sobriety testto determine whether the young person may be over the 0.08 limit or under the influenceof drugs may be appropriate. But the results of the test should not influence the zerotolerance citation because that decision should be based on evidence of drinking and noton impairment.

PER SE LAWS VERSUS BEHAVIOR

The evaluation of a suspect’s behavior using SFSTs strongly contrasts with the evidencesupposedly required under current DUI driving laws. These are based on statutes thatmake a given BAC illegal per se. In principle, the driver’s behavior is irrelevant under aper se law, which specifies that having a BAC at or above a specific level is the offense.Nominally, courts should not permit testimony regarding the appearance and behavior ofthe suspect since this is not relevant when a BAC is available. Of the 50 states havinglaws that limit the BAC while driving, only Massachusetts does not have a per se law.

It is ironic that in our legislatures we debate what level of BAC should be illegal,and we base our public information programs on the number of drinks required to arrive at

Voas F-15

a given BAC. Yet, the process of apprehending impaired drivers in the field is almostentirely based on observation of a suspect’s behavior. Sensitive, portable, handheld breath-test equipment has been available in the United States for the past 20 years. Several stateand local police departments have been equipped with such preliminary breath testers, butthey have used the devices only at the margins of the DUI enforcement process. Behavioralobservations, no matter how skilled the observers, are subjective and have a larger errorvariance than the relatively precise measures that can be made using portable breath-testdevices. What needs to be explained is why, in the presence of this technology,enforcement of DUI continues to rely almost entirely on observations of human behavior.

LIMITED USE OF PRELIMINARY SENSORS

While the field sobriety test is the principal means by which officers determine whetherto make an impaired driving charge, many officers have been equipped with preliminarybreath testers (PBTs). However, agency policies have frequently limited the use of thesedevices. Generally, the preliminary sensor is used only after the field sobriety test. Onereason frequently given for this practice is the fear that knowledge of the BAC may bias the officer’s judgment in scoring the SFST. Many departments also have a policyagainst the use of the preliminary sensor before the field sobriety test based on theargument that the officer will focus on the alcohol impairment and will ignore thepossibility of drug impairment or become too reliant on the breath-test result and fail tocollect sufficient behavioral data to support the charge in court.

A contrasting view of the use of preliminary sensors would consider that theycould enhance the possibility of detecting drug-involved drivers. The occurrence of oddor impaired behavior in the absence of an elevated BAC should be an immediate tipoff tothe officer that he or she has a drug or medical problem on his or her hands. Whenoriginally introduced, preliminary sensors were viewed as an aid to the officers inidentifying drivers impaired by diabetic, epileptic, or heart problems rather than drinking.

The preliminary breath test needs to be given equal importance with sobrietytesting in the apprehension process, and policies for when and how it is used shouldreceive the same careful consideration and planning that have gone into the developmentof field sobriety tests.

DETECTION PROCEDURES AS A FUNCTION OF BAC

Table 1 presents an analysis of the enforcement procedures needed as a function ofprohibited BAC level. At BACs greater than 0.15, detection through traditional signs ofintoxication understandable to the general public and juries should be possible in all but themost alcohol-tolerant drinkers. At this level, the PBT principally provides the officer with ameans of distinguishing those impaired by drugs or a medical problem. It is the regionbetween 0.08 and 0.14 in which the SFSTs play the most important role in detecting andmeasuring impairment. Here the PBT plays an important role in detecting those tolerantdrinkers who show little impairment on sobriety tests. In the 0.05 to 0.08 BAC region fewbehavioral signs are available for use by the police. With a few exceptions, the PBTbecomes essential to establishing a basis for arrest, and the passive sensor can play animportant, if not essential, role in identifying drinkers. Finally, in the 0.02 range, no


TABLE 1 Requirements for DUI Enforcement Based on Offender BAC

behavioral signs should be expected. Citations must be based on circumstantial evidence oron an alcohol sensor.

TWO ENFORCEMENT SYSTEMS

While the presence of alcohol cannot be observed directly by the human senses, it can beaccurately detected and measured with specialized equipment. Australia provides anexample of primary or “chemistry-based” (Voas and Lacey, 1990) DUI enforcement.Drivers can be stopped at random and when stopped are required to provide a breath test.In addition to that primary system, Australia also employs a secondary or “behavior-based” system, in which the investigation is triggered by a driver’s error resulting in acrash or in being stopped for a traffic offense. In such cases, the officer may require abreath test at his discretion. While their program involves the use of both of theseprocedures, there is strong evidence that the primary system has produced the greatestreduction in alcohol-related crashes (Homel, 1988).

Like Australia, all but one of the U.S. states have per se laws that could permitprimary or chemistry-based enforcement. In practice, we have only a secondary system inwhich two findings are required before a breath test can be used. There must be probablecause to stop a vehicle (a traffic offense or suspicious driving). Once stopped, there mustbe a basis for suspecting the driver is impaired even before the PBT can be used. Thereare, however, enforcement techniques that potentially provide the United States with aprimary enforcement system like that in Australia. Random stopping (sobrietycheckpoints) is legal in most states, and passive sensors, which do not constitute a searchunder the Fourth Amendment (Fields and Henricko, 1986), can be used to check everymotorist stopped. The use of checkpoints has been shown to be effective (Levy, Shea,and Asch, 1989; Stuster and Blowers, 1995), and the use of passive sensors has beenshown to increase the detection of over-the-limit motorists (Ferguson et al., 1995; Lundand Jones, 1987; Williams and Lund, 1984).

In the United States we have a need for both a primary and a secondary DUIenforcement system. We still need a secondary system to arrest dangerous drivers—speeders, reckless operators—and drivers in crashes, based on probable cause to stop thevehicle and observations of the driver’s behavior. What we must not lose sight of is thatsuch secondary systems do not detect a large segment of the DUI offenders. We can seethis happening when drivers without safety belts are not stopped in secondary law states.But the failure to detect and stop a drinking driver goes unnoticed. The SFST is apowerful tool for officers once they have stopped a vehicle and decide to invite the driver

Voas F-17

out of the car for a test. We should not allow the observed 90 percent accuracy of theSFST for those tested, which Dr. Burns reports, to obscure the fact that the greatestlimitation in the SFST is the number of high BAC drivers who are not tested because theydo not show the signs of impairment that lead to testing.

PROPOSAL FOR PRIMARY OR CHEMISTRY-BASED ENFORCEMENT IN THE UNITED STATES

Primary DUI enforcement in the United States will only be possible in those states whoseconstitutions permit checkpoints. Currently, most checkpoints in the United Statesinvolve large numbers of officers and are conducted infrequently on holidays such asNew Year’s Day and the Fourth of July. Because of the large number of officers requiredto operate checkpoints, they are expensive and administratively difficult to organize.Since passive sensors are not used, half of the over-the-limit drivers passing through thecheckpoint site are missed. When processing a large number of drivers, time is not takento check driver’s licenses and frequently safety belt nonuse is overlooked.

Based on current research, the optimal application of the chemistry-basedenforcement method would be the frequent mounting of checkpoints once per week or atleast twice per month as is done in North Carolina (Foss et al., 1997), Tennessee (Laceyet al., 1997), New Jersey (Levy et al., 1989), and Charlottesville, Virginia (Voas et al.,1985). To minimize costs, these checkpoints should be staffed by four to six officers; asmall group of officers works as well as a large group of officers (Stuster and Blowers,1995). Such operations are relatively inexpensive and can be mounted by relatively smallcommunities or by one or two communities working together. Officers should beequipped with passive sensors so that they can apprehend 80 percent, rather than 50 percent, of the potential offenders. This not only increases the efficiency of theoperation—officers at checkpoints with passive alcohol sensors will apprehend DUIs at ahigher rate per officer hour than officers on patrol (Voas et al., 1985)—but also increasesthe credibility of these operations because fewer legally impaired drivers will get throughthe checkpoints. The efficiency of these primary DUI checkpoint operations can befurther increased by checking for unlicensed drivers and for the use of safety belts. Incheckpoints where driver’s licenses are examined, many arrests are made of drivers whohave been suspended as a result of a previous DUI conviction (Voas et al., 1985). Ross(1991) among others has argued that driver’s license suspension is the most effectivemeans of reducing the recidivism of DUI offenders. Thus, enforcing this sanction byapprehending suspended drivers is an important function of checkpoints.

Currently, secondary enforcement in the United States suffers from a limitation:Drivers who are alcohol tolerant do not display relatively gross driving impairment and,therefore, go undetected. A particular problem is the failure to detect more of the drinkingdrivers involved in crashes. Using passive sensors would increase the number of impaireddrivers officers could detect at crash sites. If an officer stops a vehicle on suspicion ofoperator impairment, then he or she will usually require the SFST. However, manymotorists are stopped for reasons other than suspect impairment. In such cases, as manyas half of the impaired drivers will be missed (Taubenslag and Taubenslag, 1975).Requiring a PBT on everyone stopped by an officer would apprehend such offenders whoare currently missed. The British Road Safety Act (which required a PBT for drivers in


crashes, drivers committing traffic violations, or drivers suspected of drinking)demonstrated the effectiveness of this procedure for reducing traffic crashes (Ross, 1984).Two states, Kansas and Nebraska, have legislation that permits the officer to require abreath test of a driver who commits a driving offense such as going through a red light orspeeding. Behavioral-based enforcement in the United States could be significantlyimproved if states passed laws authorizing officers to use a PBT on all individuals whoare involved in crashes or who commit traffic offenses. An evaluation of the extent towhich these laws in Nebraska and Kansas have resulted in more effective DUIenforcement should be conducted.

REFERENCES


Fields, M., and A. R. Henricko. Passive Alcohol Sensors—Constitutional Implications. The Prosecutor,Vol. 20, No. 1, 1986, pp. 45–50.

Foss, R. D., D. J. Beirness, W. G. Tolbert, J. K. Wells, and A. F. Williams. Effect of an Intensive SobrietyCheckpoint Program on Drinking-Driving in North Carolina. In Proceedings of the 14th InternationalConference on Alcohol, Drugs, and Traffic Safety, Vol. 2, 1997, pp. 943–948. CERMT, Annecy, France.

Homel, R. Policing and punishing the drinking driver. A study of general and specific deterrence. Springer-Verlag, New York, 1988.

Lacey, J. H., R. K. Jones, and J. C. Fell. The Effectiveness of the “Checkpoint Tennessee” Program. InProceedings of the 14th International Conference on Alcohol, Drugs, and Traffic Safety, Vol. 2.CERMT, Annecy, France, 1997, pp. 969–975.

Levy, D., D. Shea, and P. Asch. Traffic Safety Effects of Sobriety Checkpoints and Other Local DWUPrograms in New Jersey. American Journal of Public Health, Vol. 79, No. 3, 1989, pp. 291–293.

Lund, A. K., and I. S. Jones. Detection of Impaired Drivers with a Passive Alcohol Sensor. In Proceedingsof the 10th International Conference on Alcohol, Drugs, and Traffic Safety, Elsevier Science,Amsterdam, the Netherlands. 1987, pp. 379–382.

McKnight, A. J., E. A. Langston, A. S. McKnight, and J. E. Lange. Sobriety Tests for Low Blood AlcoholLevels, under review.

Mothers Against Drunk Driving (MADD), Advocates for Highway and Auto Safety, and NationwideInsurance. Rating the States. A Report Card on the Nation’s Attention to the Problem of Alcohol- andOther Drug-Impaired Driving. Washington, D.C., 1996.

National Highway Traffic Safety Administration. Alcohol Limits for Drivers: A Report on the Effects ofAlcohol and Expected Institutional Responses to New Limits. Report DOT HS-807-692. NHTSA, U.S. Department of Transportation, 1991.

Preusser, D. F., R. G. Ulmer, and C. W. Preusser. Obstacles to Enforcement of Youthful (Under 21)Impaired Driving. Final Report DOT HS-807-878. NHTSA, U.S. Department of Transportation, 1992.

Ross, H. L. Deterring the Drinking Driver: Legal Policy and Social Control (2nd ed.). Lexington Books,Lexington, Mass., 1984.

Ross, H. L. License Deprivation as a Drunk-Driver Sanction. Alcohol, Drugs, and Driving, Vol. 7, No. 1,1991, pp. 63–69.

Stuster, J. W., and M. A. Blowers. Experimental Evaluation of Sobriety Checkpoint Programs. ContractNo. DTNH22-91-C-07204. NHTSA, U.S. Department of Transportation, 1995.

Voas F-19

Taubenslag, W. N., and M. J. Taubenslag. Selective Traffic Enforcement Program (STEP). NHTSA, U.S. Department of Transportation, 1975.

Voas, R. B., and J. H. Lacey. Drunk Driving Enforcement, Adjudication, and Sanctions in the UnitedStates. In Drinking and Driving: Advances in Research and Prevention, (R. J. Wilson and R. E. Mann,eds.) The Guilford Press, New York, 1990, pp. 116–158.

Voas, R. B., A. E. Rhodenizer, and C. Lynn. Evaluation of Charlottesville Checkpoint Operations. FinalReport under DOT Contract DTNH-22-83-C-05088. NHTSA, U.S. Department of Transportation, 1985.

Williams, A. F., and A. K. Lund. Deterrent Effects of Roadblocks on Drinking and Driving. Traffic SafetyEvaluation Research Review, Vol. 3, 1984, No. 6, pp. 7–18.

Comments on “Identification of Impairment Outside the Vehicle”

JAMES HEDLUND

Highway Safety North

F-20

INTRODUCTION

You have stopped a driver for a traffic violation or at a checkpoint. At the car window,you find enough suspicion of impairment to request that the driver step outside the car forfurther examination. What methods should you, a police officer, use to decide whether ornot to arrest the driver for impaired driving? You have two basic strategies, which can beused separately or together. Behavioral methods observe and record the driver as heanswers your questions and performs simple actions that you request. The usualprocedure is to follow the Standardized Field Sobriety Tests (SFST). Chemical methodsdirectly measure the alcohol in the driver’s breath to estimate his blood alcoholconcentration (BAC), most commonly with a preliminary breath test (PBT) device.

Marcy Burns and Bob Voas have given excellent discussions of behavioral (SFST)and chemical (PBT) strategies, respectively. In a brief discussion I briefly summarize thestrengths and weaknesses of these strategies, discuss how each fits into the framework ofstate impaired driving laws, provide my conclusions and recommendation on how each isbest used, and point out the research needs implied by these conclusions.

BEHAVIORAL TESTS: SFST

SFST research, development, and use are discussed thoroughly in Marcy Burns’ paper.To summarize very briefly:

• The SFST consist of three behavioral tests: walk and turn, one-leg stand, andhorizontal gaze nystagmus (HGN).

• The SFST were developed to determine whether a driver’s BAC exceeded 0.10 percent, the alcohol level most common in state impaired driving laws. Recentresearch has concluded that the SFST are equally valid at 0.08 percent BAC. Withmodified scoring, the SFST can be used at lower BACs.

• Field validation studies in California, Colorado, and Florida find that experiencedpolice officers using the SFST typically have fewer than 10 percent false positives (driverswho fail the SFST but whose BAC level is less than 0.10 percent) and fewer than 20 percentfalse negatives (drivers who pass the SFST but whose BAC level is more than 0.10 percent).

• The SFST are used in all states and are taught routinely to patrol officers in allstate police academies. I believe the most important issues presented by the SFST are:

– Accuracy: Can the SFST be improved by using additional tests or by modifyingthe three current tests?

– Consistency: How to assure that the approximately 400,000 patrol officers inthe United States are proficient in using the SFST.

– Legal challenges: How to assure that the SFST, particularly HGN, are acceptedin court.

Hedlund F-21

CHEMICAL TESTS: PBT

Alcohol does not affect every person identically, so behavioral tests never can estimateBAC levels completely accurately. Why not skip behavior and measure a driver’s breathalcohol directly? PBTs were developed precisely to do this. Several models are availablecommercially. They cost between $440 and $745 apiece, are quite accurate (definitelymore accurate than the SFST), and are generally reliable. Manufacturers estimate thatapproximately 100,000 PBTs currently are in use (Century Council, 1997).Approximately half the states currently allow PBT evidence to be admitted in court,typically to establish that a police officer had probable cause to arrest a driver but not toestablish the driver’s BAC level (Fields, 1998).

PBTs present their own set of issues. I believe the most important are

• Cost: providing one to each patrol officer in a jurisdiction costs at least $445 perofficer, at a time when police agencies are chronically short of funds.

• Reliability: the PBT must work every time, even under difficult conditions at theroadside (hot or cold, wet or dry, after heavy use or after sitting unused for a long time).

• Acceptance by police: some officers are reluctant to use PBTs because they trusttheir own experience more than a machine or do not want to bother with another piece ofequipment.

• Legal issues: courts must accept PBT evidence as part of the impaired drivingarrest procedure.

WHICH METHOD TO CHOOSE?

Recall again your task: to decide whether or not to arrest the driver for impaired driving.If you arrest, you want a solid case leading to conviction. If your only issue were to findthe best scientific evidence, your choice would be easy: PBTs are more accurate thanSFST. But scientific evidence is not the same as legal evidence. To understand why, letus look briefly at impaired driving laws and court procedures.

IMPAIRED DRIVING LAWS: BEHAVIORAL VS. CHEMICAL

Under a “behavioral” law it is illegal to drive while impaired by alcohol or another drug.All states have such a law. A driver’s BAC is only one piece of evidence used to establishimpairment by alcohol. Under a “chemical” or per se law it is illegal to drive with a BACexceeding a specified level. All states have a per se law for drivers under 21 with a BAClevel of 0.02 or less, and all states except Massachusetts and South Carolina have a per selaw for drivers age 21 and above at a BAC level of either 0.08 or 0.10.

In theory, PBT evidence alone should be sufficient to arrest and convict a driverunder a per se law (of course, after the officer has sufficient justification to stop the driverin the first place and then to ask the driver to step out of his vehicle). In practice, it doesnot work this way. Courts in the United States are not willing to convict solely on thebasis of chemical BAC evidence. At each step the arresting officer must collectbehavioral evidence that the driver is indeed impaired. The evidence must bescientifically rigorous enough to withstand cross-examination. It also must be reasonableenough to convince a jury that it is related to driving impairment. The SFST were


developed precisely to meet this need for standardized, scientifically based procedures.The growth of per se laws has not changed this need.

Other countries’ laws and practices differ. In Australia, Canada, and Sweden, forexample, per se laws require only chemical evidence, and behavioral methods such asSFST simply are neither needed nor used. But in the United States, some form of SFSTwill be required for the foreseeable future.

Administrative license revocation (ALR) laws are one way in which many stateshave adapted to this legal requirement. Under an ALR law, the state motor vehicleadministration suspends or revokes a driver’s license if the driver’s BAC exceeds thestate’s legal limit. It is a true chemical law. But the sanction is strictly administrative, notcriminal. If the state wishes to prosecute the driver under its criminal impaired driving orper se laws, then behavioral evidence is needed.

The role of HGN in the SFST illustrates the relation between behavioral andchemical evidence in the United States. HGN measures involuntary jerking movements of aperson’s eye. HGN is highly correlated with alcohol impairment; in fact, HGN contributesalmost all the SFST’s predictive power. But HGN is the portion of the SFST that ischallenged most frequently in court and has been disallowed by some courts. I believe thisis because HGN has no apparent “face valid” connection to driving. It is not something theaverage judge or juror can understand directly, but rather appears to be “black magic” thatmust be accepted on the basis of testimony from expert scientists. HGN is even moreindirect (or chemical) than BAC test evidence: judges and jurors know that drivingperformance decreases as BAC increases, but they may not know just from his BAC levelthat the arrested driver is sufficiently impaired that he is a danger on the road.

CONCLUSIONS AND RESEARCH NEEDS

I believe that behavioral evidence is here to stay in U.S. courts and that the SFST are criticalto provide consistent and scientifically valid behavioral evidence. PBTs can be useful tosupplement, but not to replace, the SFST. Thus the research and program challenges are tomake the SFST used in the field as good as they can be. We do not need research to improvethe SFST: Recent research has reviewed the tests, validated them at various BAC levels, andfailed to find any improved tests. While HGN does most of the work, the other two tests adda bit to the predictive power and also help considerably in court.

I see two research and program needs: (1) Improve SFST abilities andstandardization across 400,000 patrol officers through initial training and in-servicerefresher training and (2) continue to win acceptance for the SFST in court.

These needs are not traditional laboratory research; rather, they require things likedistance learning and motivation on the one hand and legal strategy development andinformation transfer on the other hand. But they are critical to the continued use andacceptance of the SFST, which in turn are critical to the entire impaired driving arrest,conviction, and deterrence process.

REFERENCES

Century Council. Combating Hardcore Drunk Driving: A Sourcebook of Promising Strategies, Laws andPrograms. The Century Council, Los Angeles, Calif., 1997.

Fields, M. Legal and Constitutional Issues Related to Detection, in this circular.

Evidence of Alcohol and DrugImpairment Obtained After Arrest

STEVE SIMON

University of Minnesota

Evidence of alcohol or drug impairment is an essential element of a criminal prosecutionfor driving while impaired (DWI) by alcohol or drugs. Such evidence, or the refusal toprovide such evidence, also is essential in the administrative license sanction area.Evidence of impairment includes:

1. Chemical testing for alcohol or drugs (obtaining and analyzing blood, breath, orurine samples);

2. Observations of the driver including:a. Physical condition, appearance, behavior, andb. Physiological and behavioral tests indicating possible impairment; and

3. Statements by the driver admitting alcohol and or drug consumption and impairment.

Chemical testing for the presence and amount of alcohol or drugs in a driver’sbody is the most important type of evidence of impairment. However, when a driverrefuses to provide such evidence, the prosecution of an individual for DWI is oftenlimited to the other types of impairment evidence listed above. In this paper I will discusspost-arrest impediments to obtaining evidence of impairment of a driver arrested fordriving while under the influence of alcohol or drugs, the dynamics and incidence ofchemical test refusal, procedures for reducing the refusal rate or reducing theeffectiveness of refusal in preventing a conviction for DWI, and the problems andpossible solutions to the issue of obtaining chemical tests for impairment from injureddrivers receiving treatment at emergency medical facilities after involvement in motorvehicle crashes.

IMPAIRMENT EVIDENCE BASED ON OFFICER’S OBSERVATION AND DRIVER’S ADMISSIONS

Officer’s Observations of Indication of Impairment

An officer’s observations of the indicia of alcohol or drug impairment corroborate anychemical tests of the amount of alcohol or drugs in a driver’s body and, in the absence ofsuch chemical tests, will be the only evidence (other than a driver’s admissions) the statehas to prove the charge of DWI. Standardized field sobriety tests have been developedthat correlate highly with alcohol impairment (Burns, 1985). Such tests can beadministered pre- or post-arrest. If obtained pre-arrest, they are used to establish probablecause to arrest as well as for substantive evidence of impairment. Such tests arevoluntary, so there are no criminal or administrative consequences for refusing to submitto them. Generally, field sobriety tests are deemed to be non-testimonial, andconsequently the driver does not have to be informed of his or her Fifth Amendment

G-1

G-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

rights to refuse such tests (Jaeckle, 1991; Pennsylvania v. Muniz, 1990). The court inMuniz did hold that the driver must be advised of his or her Fifth Amendment rights toremain silent if the driver is asked to make statements as part of the field sobriety test.The manner in which such tests are requested may increase the likelihood of the driver’sagreeing to submit to them. See the refusal section of this paper and the discussion ofmethods of requesting implied consent tests. Field sobriety tests are frequently attackedin court based on the argument that the officer’s interpretation of the performance of thetest is subjective. One way to minimize this challenge is to videotape the performance ofthe field sobriety tests (Kuboviak and Quarles, 1998). There has been an increased use ofvideotaping by officers arresting individuals for DWI. The video camera is mounted inthe police vehicle and records the driving conduct of the suspect, as well as the behaviorsof the officer and the driver (including the driver’s statements) after the vehicle has beenstopped. Another challenge to field sobriety tests is that they are conducted on rough,uneven road or shoulder surfaces in poor lighting. Videotaping the field sobriety testsconducted at the scene of the stop could reduce or eliminate these challenges.

There is concern in some jurisdictions that videotaping can be a tool for thedefendant if the videotape shows a driver who exhibits no signs of impairment whileperforming field sobriety tests. If such driver has submitted to an implied consent test,then the prosecutor should be prepared to call a toxicologist to present testimony at thetrial about the development of tolerance by heavy drinkers. Frequently such driversrefuse testing because they are repeat offenders and know that they will test very high.Criminalizing implied consent test refusals, or allowing evidence of refusal as evidencein a criminal prosecution and adopting plate impoundment for repeat offenders andrefusers, mitigates the effective defense of refusing an implied consent test and appearingnonimpaired on a videotape of field sobriety tests (see refusal section in this paper).

The admissibility of a driver’s refusal to perform such tests is determined by caselaw in each state. For example, in Florida the courts have held that a driver’s refusal toperform such tests was admissible in his criminal trial for DWI (State v. Taylor, 1995). Thecourt held that such tests are not compelled and consequently not in violation of the driver’sFifth Amendment rights. The court further held that refusal to perform such tests can be anindication of a consciousness of guilt. Compare the Florida decision here with State v.Whitehead, 1990, and Pennsylvania v. Muniz, 1990, which held that interrogation questionsnot related to informing a driver of his or her implied consent rights, but that are askedduring that process, are not admissible unless they are preceded by a Miranda warning.

Even if a driver refuses to submit to field sobriety tests, an officer can and shouldrecord observations of a driver’s physical demeanor, speech, and behavior. The driverwill be retrieving his or her driver’s license from a wallet or purse, walking from his orher vehicle to a police vehicle and subsequently to a booking facility, and answeringquestions about identity and responding to a request to submit to an implied consent test.With proper training an officer can identify and record many indicia of alcohol or drugimpairment by observing these behaviors.

Statements of Drivers Regarding Their Alcohol Consumption and Impairment

A driver’s statements can furnish very important and inculpatory information about theamount of alcohol or drugs they have consumed and the degree to which that

Simon G-3

consumption has affected their driving. The states vary in regard to whether a driver whois not formally under arrest for DWI, but who is being questioned in a police vehicle atthe scene of a traffic stop, must be informed of his or her Fifth Amendment right toremain silent. A given state should evaluate the decisions of its appellate courts in thisarea and design an interrogation format that is consistent with its courts’ decisions. Aftera driver has been arrested and taken to a law enforcement facility, the U.S. SupremeCourt requires that a driver must be informed of his or her Fifth Amendment right not toanswer such questions. However, many drivers, even after being informed of their rightto remain silent, respond to such interrogation and furnish very inculpatory information(Cassell, 1996). Officers should be encouraged to interrogate suspected impaired drivers,using a series of prepared questions about alcohol consumption, the effect of suchconsumption on the driver’s ability to operate a vehicle, and the existence of any physicalor medical condition that would explain the failure to perform field sobriety tests orprovide an adequate sample for a breath or urine test.

INVOLUNTARY CHEMICAL TESTS

Chemical tests for the presence and amount of alcohol and or drugs are usually obtainedthrough an implied consent test request procedure. The typical implied consent law allows adriver to refuse such a test. However, in many, if not all, states, if the implied consent law isnot invoked, a police officer can obtain an involuntary blood sample that can be analyzedfor the presence and amount of alcohol or drugs. This type of test is typically obtained inserious motor vehicle crash cases, in which the investigating officer determines that a driverinvolved in such crash may have consumed alcohol or drugs prior to the crash. Theinvestigating officer does not want to request an implied consent test because to do sowould give the driver, involved in a serious injury-producing crash, the opportunity torefuse to submit to an evidentiary chemical test. The results of such involuntary tests areused to prosecute these drivers for felony-level DWI offenses. Taking an involuntary bloodsample in the above situation is a seizure and as such is subject to the search and seizurerestrictions of state and federal constitutional law. In Schmerber v. California (1966) theU.S. Supreme Court upheld this type of seizure, even in the absence of a warrant. The onlylimitations on the admissibility of such tests are that the officer obtaining the test must beable to articulate the probable cause basis for seeking such a test and, in most cases, thedriver must be under arrest and the blood sample must be drawn consistent with forensicpractices. These include the use of a non-alcohol based skin cleaning solution, preservationof the sample to prevent fermentation, and proper maintenance of the chain of custody ofthe sample (Van Berkom, 1991). One criticism of obtaining this type of test is that, in theabsence of an implied consent test request, the driver’s license is subject to revocation onlyupon conviction of the criminal charge. This often outrages victims of these crashesbecause they observe the driver continuing to drive while the criminal case is pending.Minnesota has addressed this issue by amending its implied consent law to authorize aninvoluntary blood test after a driver has refused an implied consent test if the police officerhas probable cause to believe the driver was involved in an injury-producing crash (seeMinnesota Statute 169.123, Subdivision 4 (a)). Pursuant to this law a police officer at acrash scene can request an implied consent test and, if the driver refuses, issue a notice ofrevocation and then order an involuntary blood test, the result of which can be used toprosecute the driver criminally for the felony-level DWI crime.


Several states allow a Schmerber type of blood sample seizure even in non-accident DWI cases (see State v. Sauer, 1993, involuntary test allowed in noninjury DWIsituation; State v. Slaney, 1995, involuntary test prohibited by statute in noninjury DWIsituation). Some law enforcement agencies obtain a Schmerber involuntary blood testwhen they arrest a repeat DWI offender, even if there was no crash. The reasoning is thatrepeat DWI offenders frequently refuse to take an implied consent test. The typicalpenalty for refusing an implied consent test is license revocation. However, repeat DWIoffenders frequently are currently revoked at the time of their repeat DWI arrest, so thethreat of a license revocation is meaningless to them.

The repeat DWI offender often has developed a tolerance for alcohol andconsequently exhibits fewer of the external signs of alcohol impairment than a non-chemically dependent driver. The repeat DWI offender also frequently refuses to submitto field sobriety tests. In this situation the state has little evidence of impairment withwhich to prosecute the repeat DWI offender. Law enforcement agencies, recognizing thedifficulty of prosecuting this type of repeat offender, have more incentive to obtain aSchmerber involuntary blood test, the analysis of which is admissible in the prosecutionof the repeat DWI offender. One possible challenge to the admissibility of an involuntaryblood test taken from a repeat DWI offender would be that taking such a test only fromrepeat DWI offenders is a violation of state and federal constitutional provisionsguaranteeing equal protection of the law. Law enforcement agencies should work withtheir prosecutors to develop a protocol that would allow them to obtain this type ofinvoluntary chemical test for repeat DWI offenders. (See refusal section for a discussionof sanctions for refusal in addition to license revocation.)

VOLUNTARY NONIMPLIED CONSENT TESTS

In many states a police officer can obtain a voluntary non-implied consent evidentiarytest from a suspected DWI driver (see State v. Slaney, 1995). A voluntary test involves apolice officer asking a DWI driver if he or she will take a breath, blood, or urine test. Nomention is made of any license revocation consequences for refusing to submit to such atest, and in fact there are none. If the driver refuses to take a voluntary test, then theofficer may take an involuntary blood sample if the jurisdiction allows one. Voluntarytest request situations arise when an officer does not have sufficient facts to request animplied consent test, and the driver has not been involved in an injury-producing crash.Police officers should be educated about the admissibility of such voluntary tests in theirjurisdiction. Voluntary testing could, in some situations, be the only way in which anevidentiary test can be obtained.

EVIDENTIARY TESTS

In a criminal prosecution or administrative license revocation procedure for a DWIviolation, the admissibility of an evidentiary test for the presence and amount of alcoholor a controlled substance is governed by the statutes, court decisions, and administrativeregulation of each state (DWI evidentiary testing regulatory framework). In this section Iwill identify the numerous issues that have been raised in challenges to the admissibilityof such tests. The focus will be on issues that arise in the test request and administration

Simon G-5

process. Also discussed will be certain legal concepts and forensic principles that are partof many states’ DWI evidentiary testing regulatory framework that impede or complicatethe admission of evidentiary tests.

For a concise yet thorough discussion of the dynamics of chemical testing,including reference to some of the core issues about which states must make statutorypolicy decisions, see Dr. Lowell Van Berkom’s “Chemical Test Evidence in DWI Cases:Some Issues and Challenges.” Every year throughout the nation court challenges to theaccuracy of alcohol analysis number in the thousands, and many result in reversalsbecause of the “failure of law enforcement agencies to adhere to the established technicalor administrative procedures or to the necessary scientific safeguards” (Dubowski, 1986).Technical procedures are based on principles of testing that require testing to be accurateand reliable. Consequently, they cannot be streamlined very much, other than through thedevelopment of new or improved testing technology. Administrative procedures andstatutory regulations can often be streamlined to reduce the time and complexity involvedin obtaining an evidentiary test. There are two periodicals that collect and summarizecourt decisions involving all aspects of DWI. Included in these periodicals are decisionsthat address forensic and procedural issues regarding the admissibility of evidentiary testresults in DWI criminal and administrative cases. Policy makers and researchers shouldrefer to these journals on a regular basis to remain current on the significant issues relatedto evidentiary test admissibility. These two periodicals are the “Drinking Driving LawLetter,” published by Clark Boardman and Callaghan, and the “Traffic Law Reports,”published by Knehans-Miller Publications. The “Drinking Driving Law Letter” alsoincludes short summaries of research articles in the broad area of alcohol, drugs, anddriving, including chemical testing and the physiology of alcohol and alcoholism. Areview of these periodicals indicates that the following issues are frequently raised in theadmissibility of evidentiary testing in criminal and civil DWI cases.

Test Request Procedure

• Challenges to sufficiency of the grounds to request an implied consent test.• Improper advising of implied consent rights and obligations.• Confusion on part of driver regarding test submission and right to refuse.• Confusion between implied consent rights and Miranda rights.• Language barriers to understanding implied consent advice.• Right to attorney before decision regarding test submission.• Choice of test (see discussion below).• Officer documentation necessary for admissibility of test.

Breath Tests

• Availability of testing devices.• Availability of qualified operators.• Observation period.• Administrative regulation; non-compliance.• Statutory regulation; non-compliance.• Instrument maintenance, repair, calibration, and certification.


• Operator training and certification.• Test instrument issues: instrument malfunction; deficient sample; substance in

mouth; room air contamination; contaminated sample; radio frequency interference;margin of error; computer printer error; maintenance and repair defects; maintenance andrepair records; lack of specificity of instrument for alcohol; correlation factor; calibrationtests: simulator solution defects, age, and preparation certification; and failure to savebreath sample.

• Inability to provide breath sample (duty to detect; and duty to advise officer ofinability).

Blood Tests

• Sample obtaining procedure (forensic test kits vs. medical sample; identity andnature of solution used to cleanse skin; training and qualification of blood drawer; andinforming driver of blood drawer qualifications).

• Sample preservation, handling, and transportation (including chain of custody).• Preservation and availability of sample portion for driver.• Obtaining medical records of blood test done for medical purposes after accident.• Analysis of blood sample for drugs when only alcohol test requested.• Interpretation of results (results given in wrong unit of measurement, weight, or

volume; required presence of blood drawer at hearing/trial; required presence of bloodanalyzer at hearing/trial; and admissibility of sample draw procedure and analysisdocumentation).

Urine Tests

• Procedure for obtaining valid sample.• Provision of test sample in private.• Driver adulteration/contamination of sample.• Sample preservation, handling, and transportation (including chain of custody).• Preservation and availability of sample portion for driver.• Medical condition; inability to give sample.• Required presence of urine sample observer at hearing/trial.• Required presence of urine analyzer at hearing/trial.• Analysis of urine sample for drugs when only alcohol test requested.• Interpretation of results (results given in wrong unit of measurement, weight, or

volume; required presence of blood drawer at hearing/trial; required presence of urineanalyzer at hearing/trial; and admissibility of sample draw procedure and analysisdocumentation).

Four Areas Generating Frequent Challenges to Admissibility of Evidentiary Test Results

Urine/Blood, Breath/Blood Conversion Ratio

In states that prohibit operation of a vehicle with a certain blood alcohol level, breath andurine test results must be converted to a blood alcohol concentration. The conversion

Simon G-7

factor used in many states for breath to blood is 1 to 2100. That means that for 1 unit ofalcohol in a person’s breath, there are 2100 units of alcohol in the person’s blood. Thisconversion factor is an average of a range between 1 to 1150 and 1 to 3000 (Dubowski,1986). The current conversion ratio is now thought to be 1 to 2300 (Dubowski, 1986). Anindividual driver’s conversion ratio can be anywhere within the above range and,consequently, the use of the 1 to 2100 conversion ratio has generated numerousevidentiary challenges to its accuracy in a given criminal prosecution. This argument isparticularly effective in cases involving a driver with a low alcohol concentration. Asimple solution to this problem, and one that many states have adopted, is to prohibit theoperation of a motor vehicle with a certain amount of alcohol in a driver’s blood, breathor urine. By including breath and urine as a prohibited condition, conversion is no longernecessary or even relevant (Dubowski, 1986, Van Berkom, 1991). See also People v.Ireland (1995), wherein California’s adoption of this statutory framework (blood, breath,or urine) was challenged and upheld. In Ireland the court acknowledged the tremendousnumber of hearings and amount of time taken up by court cases involving this issue.

Relation of Alcohol Concentration at Time of Test to Time of Driving

A substantial number of states require the prosecution to introduce evidence of thealcohol concentration (breath, blood, or urine) of the driver at the time of driving, eventhough the driver’s alcohol concentration was measured some time after driving. This isdifficult if not impossible to prove in many DWI prosecutions (Van Berkom, 1991). Instates that require this “relation back” evidence, drivers frequently argue that theiralcohol concentration was below the state’s legal limit at the time of driving even thoughtheir alcohol concentration was over the legal limit at the time of testing. If the driver’salcohol concentration was close to the legal limit at the time of testing, then this argumentcan be persuasive. The relation back requirement can and does result in acquittals orreversals. A simple solution to this impediment to convicting the impaired driver is theadoption of a DWI law that prohibits having an alcohol concentration over the state’slegal limit within a certain time of driving (Van Berkom, 1991). Minnesota’s DWI law,Minnesota Statute 169.121 Subdivision 1 (e) and (f), includes this prohibition, (0.10 ormore within two hours of driving), and it has effectively eliminated this issue from DWIprosecutions.

Driver’s Right to an Independent Test

Many states’ DWI laws allow a driver, while in custody, to obtain an “independentalcohol concentration test” after submitting to the state’s evidentiary alcoholconcentration test. A driver also has a constitutional right to obtain evidence of anexculpatory nature (California v. Tombetta, 1984). It is important to note that when lawenforcement takes an involuntary blood test for any type of criminal investigation,including DWI, the suspect/defendant is not entitled to an independent test while incustody. The statutory right to an in-custody independent test was adopted in the earlydays of breath testing, when the breath-test devices were not perceived to be as accurateor reliable and could be manipulated such that the instrument could be made to read anybreath alcohol level. Police-officer breath testing removed alcohol concentration testing


from the laboratory, where it was performed by medically trained personnel. There was alack of trust in the accuracy of a forensic test procedure that was conducted by a non-scientist “layman” police officer. Consequently, an independent test right allowed thedriver to obtain a traditional laboratory or medical-type alcohol concentration test thatcould be used for protection against the then-new non-laboratory testing that hadsignificant criminal consequences. Currently, law enforcement-operated breath-testinginstrumentation is highly reliable, accurate, and tamperproof (Dubowski, 1986). Evenwith the evolution and improvement of breath-testing devices few states have eliminateda driver’s right to an independent test. It appears that there is an increase in the demandfor such tests in DWI cases, because it imposes additional duties on the arresting officer.The right to an independent test while in custody imposes complex burdens and issues onlaw enforcement officials. These include, but are not limited to, duty to advise of right,access to telephone to arrange test, duty to transport driver for test, access to the custodyfacility and driver by medical or laypersons, and duty to pay for an indigent driver’sindependent test. A review of the appellate decisions in this area indicates thatevidentiary tests are frequently suppressed because the police somehow interfered withthe driver’s right to obtain an independent test. The need for an independent test is nolonger necessary in light of the reliability, accuracy, and tamperproof nature of currentbreath testing. Portions of evidentiary blood and urine samples have been made availableto drivers for many years. Technology exists that captures and preserves a portion of thedriver’s breath during the state’s evidentiary breath test (Dubowski, 1986). Theavailability of such samples could mitigate a driver’s need for an independent test. It maybe more time- and cost-effective for a state to adopt a breath, blood, and urine samplepreservation program and to eliminate a driver’s statutory right to an independent test.This area needs further investigation and policy analysis.

Choice of Evidentiary Test Method

Some states require that the driver be given a choice of the type of implied consentevidentiary test he or she will take. This choice was adopted in the early period of impliedconsent laws and was based on two reasons. The first reason for allowing drivers this choicewas that early breath-test devices were not as reliable, accurate, or tamperproof as currenttesting instruments. Also, as previously discussed in the section on the right to anindependent test, non-laboratory layman alcohol testing was a new development in forensicevidence gathering and consequently was viewed with some skepticism by policy makersand the public. So, to protect the rights of driver/defendants who were requested to take atest, they were given the right to choose the type of test they would take. A second reasonfor allowing drivers this choice was that many people have a fear of needles—having aneedle inserted into their arm is a psychologically and physically painful experience. In theearly period of implied consent evidentiary testing, a blood test was perceived to be moreaccurate than the breath-test technology then in use. While the fear of needles continues tobe present in many drivers, breath-testing technology is now very accurate, reliable, andtamperproof (Dubowski, 1986). Allowing a driver to choose the type of test he or she willtake creates additional issues for law enforcement. These include, but are not limited to,increased time between driving and testing if a blood test is chosen (although this can beattenuated by the adoption of a DWI law that prohibits driving within a certain period of

Simon G-9

testing, (see discussion of “relation back” evidence); the blood drawer and analyzer (if ablood test is chosen) become additional witnesses that the state must produce at trial;inaccuracy of urine tests, which create additional areas of challenge to drivers; cost to thestate of drawing a blood sample; delay in obtaining sample analysis if a blood or urine testis chosen; and confusion and error on the part of the officer in properly informing of right tochoice of test (resulting in evidentiary test suppression or reasonable refusal).

States that currently allow a choice of test should follow the lead of those statesthat have modified their DWI law so that the officer now chooses the test (which in mostcases will be a breath test), and only if the officer chooses blood or urine does the driverhave a choice of an alternative test.

IMPLIED CONSENT TEST REFUSAL

Implied consent test refusal is a serious problem in that refusers are significantly lesslikely to be convicted of DWI (Ross et al., 1995). Aggravating that problem is that refusalrates appear to increase with offense level (Ross et al., 1995). In all but a few states theonly consequence for test refusal is license revocation. However, for a repeat offenderwhose license is already revoked, this consequence is meaningless. In this section I willdiscuss the effect of refusal on conviction rates, why drivers refuse, possible methods thatcan be implemented to reduce the rate of refusal, and which laws a state can adopt toreduce the effectiveness of refusing an implied consent test. I will not discuss what isdeemed a refusal as defined by the statutes and courts of each state.

The rate of refusal to submit to an implied consent test varies from state to state.However, the national refusal rate appears to be approximately 20 percent, with some statesexperiencing refusal rates approaching 50 percent (Jones et al., 1991). Ross et al. (1995)examined the refusal dynamic in Minnesota in 1992. In addition to analyzing refusal andconviction data from the Minnesota Department of Public Safety, interviews wereconducted with police, prosecutors, judges, defense attorneys, probation officers, andmultiple DWI offenders. The study found that Minnesota’s overall refusal rate wasapproximately 25 percent, similar to the national average. Refusal had a significant effecton the likelihood of conviction for DWI. The study found that 75 percent of first offenderstaking an implied consent test were convicted of a DWI-level offense, and 20 percent wereconvicted of a reduced non-DWI-related charge. Conviction in almost all cases wasobtained by a guilty plea. Only 58 percent of first offenders refusing an implied consent testwere convicted of a DWI-level offense, and 31 percent were convicted of a reduced non-DWI-related charge. Repeat DWI offenders taking an implied consent test had a DWIconviction rate of 87 percent and a reduced charge conviction rate of 4 percent. RepeatDWI offenders who refused an implied consent test had a DWI conviction rate of 76 percentand a reduced charge conviction rate of 7 percent.

Ross et al. attempted to identify why drivers refuse to take an implied consenttest. They detected a significant difference in the rate of refusal between drivers arrestedby the Minnesota State Patrol, with a 7 percent refusal rate, compared with a 22 percentrefusal rate for all other police agencies in the state. State Patrol and other police officerswho devoted all of their time to traffic enforcement were able to articulate methods andtechniques they had developed to increase compliance with a request to submit to animplied consent test request. These techniques were related to reducing the fear and


antagonism that the arrested (and intoxicated or impaired) driver developed during thearrest process. Ross et al. called for further research into this dynamic. They believe that,if such techniques can be identified and taught to other law enforcement officers, refusalrates could be lowered. Ross’s research also identified other refusal dynamics that wouldnot be affected by changes in test request procedures by law enforcement officers. Thesedynamics involved drivers making a conscious choice to refuse based on the knowledgeand/or belief that test submission increases the likelihood of a conviction in their case.The preceding statistics validate this belief. An additional dynamic was identified in testrefusers who were characterized as fundamentally angry, hostile, and anti-social.

Even the adoption of more effective test request procedures will not totallyeliminate refusal. There will be a significant part of the test request population who willalways refuse. A state’s DWI, implied consent, and motor vehicle registration laws canbe changed to deal with this population. Criminalizing test refusal is a powerful andeffective tactic. Alaska, Nebraska, Minnesota, and Ohio have adopted this type of law.The laws in these states authorize incarceration upon conviction for such an offense. NewYork has criminalized the refusal to submit to a preliminary breath test, and Rhode Islandimposes a fine and “60 hours of public service” for refusing an evidentiary test. Inessence, it becomes a separate DWI crime for a driver to refuse to submit to an impliedconsent test. The criminal penalty for refusing is the same as the penalty for a convictionfor a substantive DWI offense. In addition, repeat offenders who refuse receive the sameincreased penalty for refusing as they would have received if they had been convicted ofa substantive DWI offense.

Ross et al. studied the criminalization of test refusal in Minnesota. They foundthat while refusal rates did not decrease, conviction rates did increase by about 5 percent.Their interviews with police and prosecutors indicated that both groups stronglysupported the new crime and believed that it “leveled the playing field” and was asignificant and powerful tool that increased convictions. Prosecutors and judges alsobelieved that the new crime decreased trials. This dimension was not studied by Ross et al. because there was no statewide data base that tracked the manner (trial or plea) inwhich DWI cases were resolved. It is important to note that at the time of the Ross study,the refusal crime only applied to repeat DWI offenders. Minnesota expanded the refusalcrime to first offenders in 1991. It also important to note that in a state that criminalizestest refusal, an attorney, advising a driver suspected of DWI, cannot advise that driver torefuse to take the test.

While criminalizing implied consent test refusal may be controversial in thatmany people believe it is philosophically wrong to make it a crime to refuse to give thegovernment evidence that it can then use to prosecute you, the crime is in essence nodifferent than the crime of failing to file an income tax return. It is this author’s strongopinion that states should criminalize implied test refusal, at least for repeat DWIoffenders.

An additional reason for criminalizing test refusals is that by doing so the need forlaw enforcement officers to seek a Schmerber involuntary blood test is reduced oreliminated in non-accident DWI cases. The driver can be prosecuted for the crime ofrefusal, along with the crime of DWI (no test).

Another response to test refusal that states can adopt to reduce the effectiveness ofrefusing to submit to an implied consent test is to adopt some form of vehicle

Simon G-11

immobilization for repeat DWI offenders who fail or who refuse to submit to an impliedconsent test. Repeat DWI offenders continue to drive after their license is revoked for animplied consent or DWI incident (Ross and Gonzales, 1988). Thus the increasedrevocation period for refusing an implied consent test has little effect on this group.However, if the vehicle driven by the DWI offender is immobilized because the driverrefused the test, driving a vehicle will be much more difficult. Research has shown thatplate impoundment, a form of vehicle immobilization, reduces overall DWI recidivismby approximately 25 percent, with the reduction in recidivism decreasing as the numberof prior offenses is increasing (Rodgers, 1994). Vehicle immobilization will be politicallyand philosophically easier for a state to adopt than criminalizing test refusal. In fact, thereis a significant trend among the states to adopt vehicle sanctions for repeat DWIoffenders. One form of vehicle immobilization, plate impoundment, is a low-cost, easilyimplemented vehicle sanction that does not burden law enforcement or the courts. Statesdesiring to reduce the effectiveness of implied consent test refusal should consideradopting this law.

Another response to refusal that states can adopt is to authorize, by statute, theadmission into evidence of a driver’s test refusal in the criminal trial for DWI. Juriesfrequently question why there is no chemical test in a DWI refusal criminal trial. If suchevidence is not admissible, then they are left to speculate that the police may have failedto offer one. In those cases, the driver argues to the jury that he or she was not impairedand that there is little or no evidence of impairment. If evidence of test refusal isadmissible, then the prosecutor can cross-examine the defendant about why he or sherefused, if the defendant now claims he or she was not impaired.

OBTAINING BLOOD SAMPLES FROM INJURED DRIVERS

In this section I will discuss the problem of drivers injured in DWI-related crashes whoavoid prosecution for that crime. NHTSA estimates that in 1996 there were 3,511,000traffic injuries and, of those, 321,000 involved alcohol (NHTSA, 1996). An analysis ofthe traffic and criminal records of injured drivers admitted to trauma centers who weredetermined to have consumed alcohol before or while driving indicates that very few(some studies show charging rates as low as 5 percent) were ever charged with a DWIoffense (Runge et al., 1996; ICADTS Reporter, 1996; Lillis et al., 1996). The researchindicated that the more severe the injury, the less likely the driver was to be charged.Runge and Lillis identified several factors that contributed to this low charging rate.These include the focus of the investigating officer on arranging emergency medicalservices for the injured vehicle occupants at the scene of the crash, the severity of adriver’s injuries reducing or eliminating an officer’s ability to detect signs of alcoholconsumption and impairment, emergency room personnel’s reluctance to interrupttreatment to obtain a forensic blood sample pursuant to a law enforcement request, andphysician-patient medical privilege and data privacy laws that prohibit emergency roompersonnel from contacting a law enforcement agency to notify the agency of the presenceof an impaired driver (Runge et al., 1996; ICADTS Reporter, 1996; Lillis et al., 1996).When an investigating officer detects signs of alcohol consumption on the part of a driverinjured at the scene of a motor vehicle crash, an involuntary blood sample cansubsequently be obtained at the request of that officer (see previous discussion of


involuntary blood samples). The core problem is that in crashes with severe injuries,crash scene detection of alcohol consumption is difficult if not impossible. Theinvolvement of alcohol is first detected at the trauma center where the injured driver isreceiving treatment. Alcohol consumption or impairment is detected at the trauma centerbased on the odor of alcohol, physiological or behavioral signs of impairment, or amedical blood alcohol test. Runge reports that at the Charlotte, North Carolina, traumacenter all patients are required to have a blood alcohol determination. The compliancerate with this requirement is 85 percent. At the trauma center, even if the emergencyroom personnel wanted to notify a law enforcement officer of the presence of an impaireddriver, the medical privilege data privacy laws of approximately 40 states prohibit suchcontact and disclosure (Wherry, 1996). Absent a physician-patient medical privilege,emergency room or trauma center personnel could contact law enforcement officials whocould then either request a forensic blood sample or subpoena the medical blood testresults for use in a criminal prosecution (see State v. Haselman, 1993; People v. Menssen,1994; results of blood test taken for medical purposes admissible at DWI trial).

The following steps can be taken by a state that desires to increase the prosecutionrate of alcohol-impaired injured drivers.

First, modify the state’s physician-patient medical privilege law to allow medicalpersonnel to report the presence and identity of persons, injured in motor vehicle crashes,who exhibit signs of alcohol or drug impairment or whose diagnostic blood alcohol testindicates the presence of alcohol. Three states have such laws: Pennsylvania has had sucha law for many years, and Hawaii and Illinois recently amended their physician-patientmedical privilege laws to allow disclosure. The permissive reporting statutory languageshould apply to all persons injured in car crashes, not just to drivers. Emergency roompersonnel will know the nature and cause of the person’s injuries but will not know if theinjured person was the driver. In most cases the law enforcement officer investigating thecrash will have that information, at least to a probable cause level sufficient to request aSchmerber involuntary blood test. The amendment of a state’s physician-patient medicalprivilege law to allow or to require such reporting is opposed by many members of themedical community, based on a concern that they want to be caregivers not policeofficers in regard to their patients. They are also concerned that the injured person may beless likely to seek medical treatment if the person knew that his or her treatment providercould or is required to report the person to a law enforcement agency. On the other hand,many emergency room and trauma center personnel are frustrated when they are unableto report to law enforcement a person injured in a car crash caused by that person’s ownalcohol consumption and abuse.

Second, establish procedures and communication protocols for use by emergencyroom and trauma center personnel to identify an alcohol- or drug-impaired crash victimand to notify law enforcement of the presence of such an injured individual. Having theseprocedures in place will increase the likelihood and efficiency of their use. Even with aright to notify law, absent established identification and notification procedures,emergency room and trauma center personnel may be reluctant to notify law enforcementbecause of a perception that such a process may be time-consuming, interfere with thedelivery of medical services, and require numerous court appearances.

Third, educate law enforcement in the areas of the extent of alcohol involvementin crashes and the current low rates of charging these drivers. Training could also includeimproved procedures to increase the identification of an injured driver’s alcohol

Simon G-13

impairment. The availability and use of passive alcohol sensors could increase theidentification of such drivers.

SUMMARY

Evidence of impairment is crucial in a DWI prosecution. This paper has identified areasin which problems occur in obtaining that type of evidence and also changes that can beadopted or implemented that can reduce or eliminate those problems.

Procedures for obtaining non-chemical test-type evidence of impairment need tobe developed, standardized, and taught to law enforcement. Observations of drivers, fieldsobriety tests, interrogation of drivers, and videotaping of these forms of evidence-gathering can provide prosecutors with valuable evidence of impairment, and, in caseswhere the driver refuses to submit to a chemical test, the only evidence of impairment.

Involuntary and voluntary non-implied consent chemical tests can be a viableprocess for obtaining chemical tests for impairment.

Training and education need to be increased to reduce the incidence of testoperator error that often results in the suppression of chemical tests for impairment.

Test refusal is a significant problem in this country. States can adopt statutorymeasures, including criminalization of test refusal, vehicle immobilization, and theadmission of test refusal into evidence to reduce the effectiveness of implied consent testrefusal. Research is needed in the area of the test request process to identify methods thatcan reduce the refusal rate.

In the area of evidentiary chemical testing states should modify their DWI Laws to

1. Prohibit driving with a given breath or urine or blood alcohol concentration andthus eliminate the necessity to introduce conversion evidence;

2. Allow the test-requesting officer to select the type of test the driver must take;3. Prohibit having a given alcohol concentration within two hours of the time of

driving and thus eliminate the necessity of introducing relation back evidence.4. Eliminate the right of a driver to request an independent test while in custody.

Injured impaired drivers are seldom charged with DWI. Physician-patient medicalprivilege laws should be amended to allow emergency room and trauma center personnelto report persons injured in car crashes who exhibit indicia of alcohol or drugconsumption and/or impairment to law enforcement agencies.

REFERENCES

Burns, M. Field Sobriety Tests: An Important Component of DUI Enforcement. Alcohol, Drugs, andDriving, Vol. 1, No. 3, July–September, 1985.

Cassell, P. Miranda’s Social Costs: An Empirical Assessment. Northwestern University of Law Journal,Winter 1996.

Dubowski, K. Recent Developments in Alcohol Analysis. Alcohol, Drugs, and Driving, Vol. 2, No. 2,April–June, 1986.

ICADTS Reporter. Illinois Works Toward Model Procedures and Training to Increase Arrests of InjuredImpaired Drivers. ICADTS Reporter, Vol. 7, No. 4, Fall 1996.


Jaeckle, N. Constitutional Issues in Roadside Sobriety Testing. Colorado Lawyer, January 1991.

Jones, R., H. Joksch, and C. Wiliszowiski. Implied Consent Refusal. Technical Report. National HighwaySafety Administration, Washington, D.C., 1991.

Kuboviak and Quarles. DWI Mobile Videotaping for Police and Prosecutors: Policies, Procedures andLaw. Institute of Police Technology and Management, University of North Florida, 1998.

Lillis et al. Impediments to Arrest of the Injured Drinking Driver. ICADTS Reporter, Vol. 7, No. 4, Fall 1996.

National Highway Traffic Safety Administration. Traffic Safety Facts, 1996—Overview, Alcohol. NHTSA,U.S. Department of Transportation, 1996.

Pennsylvania v. Muniz, 434 U.S. 106, 110 S. Ct. 2368 (1990).

People v. Ireland, 33 Cal.App. 4th 680 (1995).

People v. Menssen, 636 N.E.2d 1101 (Ill. App. 4 Dist. 1994).

Rodgers, A. Effect of Minnesota’s License Plate Impoundment Law on Recidivism of Multiple DWIOffenders. Alcohol, Drugs, and Driving, Vol. 10, No. 2, 1994.

Ross, H., S. Simon, J. Cleary, R. Lewis, and D. Storkamp. Causes and Consequences of Implied ConsentTest Refusal. Alcohol, Drugs, and Driving Vol. 11, No. 1, Jan.–March 1995.

Runge, J. et al., Enforcement of Drunken Driving Laws in Cases Involving Injured Intoxicated DriversAnnals of Emergency Medicine, Jan. 1996.

Schmerber v. California, 348 U.S. 757 (1966).

State v. Haselman, 635 A2d. 310 (Conn. App. 1993).

State v. Sauer, 501 NW2d. 673 (Minn. Ct. App. 1993).

State v. Slaney, 653 So.2d 422 (Fla. Ct App. 1995).

State v. Taylor, 648 SO. 2D 01 (Fla. 1995).

State v. Whitehead, 458 NW2d 145 (Minn. Ct. App. 1990).

Van Berkom, L. Chemical Test Evidence in DWI Criminal Cases: Some Issues and Challenges. Alcohol,Drugs, and Driving, Vol. 7, No. 3–4, 1991.

Wherry, E. Vampire or Dinosaur: A Time To Revisit Schmerber v. California. 19 Am. J. Trial Advoc, 503,Spring 1996.

Comments on “Evidence of Alcohol and DrugImpairment Obtained After Arrest”

HERBERT A. MOSKOWITZ

Southern California Research Institute

The 1998 TRB Committee Conference heard a variety of papers that examined theproblems and issues in detecting and convicting drivers using alcohol and other drugs. The paper by Steve Simon presented a thorough review of the chemical and/orbehavioral evidence required for successful prosecution of a DWI arrest. Successfullydeterring drivers from using toxic substances while driving requires them to believesuch behaviors will be detected and punished by law enforcement. Before discussingthe issues raised by Steve Simon about the acquisition of impairment evidence, let usfirst examine the current success rate of the legal system in prosecuting alcohol drivingarrests.

The following discussion uses data from the annual reports of the California DUIManagement system. In the latest report (Tashima and Helander, 1996), DUI convictionrates were 73 percent overall for the 231,696 arrests in 1993 and the 206,583 arrests in1994. An additional 9 percent of the 1993 arrests resulted in alcohol-related recklessdriving convictions. A further 2.4 percent of the arrests had convictions for offenses otherthan DUI or reckless driving. Thus, approximately 85 percent of the original 1993 DUIarrest population received convictions according to state records.

What occurred with the other 15 percent who were arrested but apparently notconvicted? This issue was examined in 1993 (Tashima and Helander, 1993), when arandom sample was obtained of 1,000 DUI nonconviction arrests from 1990, as reflectedin the state master file. Detailed pursuit of these non-convicted arrests determined that 37 percent had actually been convicted, but the records failed to be transmitted to thestate. Another 33 percent of the non-convicted arrests failed to appear at court. Thiswould equate to roughly 5 percent of the original arrestee population. It is believed toinclude drivers departing the state and thus avoiding subsequent court warrants. From thesample of non-convicted arrestees, only 6 percent were found not guilty, or less than 1 percent of the original arrestee population. Other reasons for non-conviction includefelony imprisonment, death, etc.

Thus, of the more than 200,000 California drivers arrested annually, less than 1 percent were found not guilty.

Participants at the TRB Conference suggested that not all states were assuccessful as California in prosecuting DUI arrestees. However, at least for California,DUI prosecution has a strikingly successful conviction rate. DUI prosecution is certainlymore successful than most prosecutions for activities such as robbery or murder. FewDUI arrestees are actually dismissed as not guilty.

I concur with Steve Simon that attention should continue to be paid to properdevelopment of evidence. But at this time in California apparently the procedures ineffect are capable of sustaining a high rate of court convictions. It would appear that forCalifornia greater emphasis should be placed on determining the reasons for the failure ofroughly 5 percent of the arrestees to appear in court.

G-15


Based on the above, I would argue that the deterrence value of DUI convictions isnot hindered by an already high conviction rate. Estimates at the conference suggest thereare still more than 500 DUI occurrences for every DUI arrest. If there is an insufficientdegree of DUI deterrence remaining in California, then the reason for the lack ofdeterrence should be sought in other areas. Incidentally, the less than 1 percent ofarrestees who managed to avoid conviction in court contests also were subjected to strongnegative reinforcements, including the cost of their attorneys.

These comments do not take away from Steve Simon’s excellent suggestion for improving the evidence gathered for DUI convictions in contested cases. Certainlythe best and simplest evidentiary evidence is the automated alcohol breath tester withits ease of usage and resistance to tampering. Because of problems of variability inobtaining behavioral evidence of impairment, passage of per se laws requiring onlybreath or blood samples represents a great contribution in improving the DUIconviction rate. Refusal to provide body fluid specimens, including breath, perhapscan be best dealt with as a separate offense whose punishment is equivalent to DUIconviction.

Steve Simon emphasizes the importance of behavioral signs of impairment beingobserved by officers, especially when body fluid samples are difficult to obtain. However,in a Southern California Research Institute study (Tharp et al., 1981) breath samples wereobtained from drivers stopped and interviewed by police officers for possible infractions.The study reported that officers released more drivers with blood alcohol concentrations(BACs) above 0.10 percent than they detected and arrested. This result is similar to theWells et al. (1997) study at random sobriety checkpoints. The results of the precedingtwo studies, as well as others, suggest that it is the problem of detecting and arrestingDUI offenders rather than the problem of convicting them that is hindering the deterrenceeffects of law enforcement.

It is not only police officers who have difficulty observing evidence of impairmentor intoxication in drivers with blood alcohol levels above the legal limit. In 1932,Widmark reported on more than 500 drivers in Sweden arrested for possible drivingunder the influence of alcohol. Sweden required these drivers to be independentlyexamined by physicians in police stations on a seven-item behavioral test battery. Only ata BAC level above 0.14 percent was it at or above 50 percent probability that a driverwould be evaluated as under the influence. It required a BAC level above 0.26 percentbefore 100 percent of drivers were declared under the influence.

Urso (1981) had patients in a Pittsburgh hospital emergency room graded byphysicians on simple behavioral tests leading to a judgment of intoxicated or notintoxicated. Just before emergency room discharge, blood samples were obtained. Thepatients considered sober who had alcohol present had a mean BAC level of 0.272 percent.The highest BAC in an individual considered not intoxicated was 0.54 percent. Thus, thefailure of police to identify behavioral evidence of intoxication is a universal problem forsubjective evaluation of alcohol influence and is not limited to any profession. Clearly, inboth Sweden and America, trained physicians were no better than police at making thedetermination. Legislation that permitted DUI convictions based on per se BAC levelswere a major step forward in increasing the rate of DUI convictions for arrestees. Itobviated the difficulty of judging behavioral signs of alcohol impairment. This statementshould not cast doubt on the efficacy of the DOT-approved sobriety test. Administered by

Moskowitz G-17

adequately trained police officers, the test produces evidence of impairment in 85 to 90 percent of individuals in the 0.08–0.10 percent range. Even greater success rates occurat higher BACs. However, one has to have a properly administered sobriety test. Simpleobservation of random behavior is not a successful substitute. Moreover, even thesebehavioral tests will never be as successful as the breath testers, given the existence of per se laws.

From a systems analysis viewpoint, given the current high efficiency of theexisting court system in convicting DUI offenders, at least in California, prime emphasisshould be placed on bringing more drivers into the court system by improving methodsfor detecting the presence of alcohol impairment.

One suggestion that would greatly enlarge the alcohol driving arrest rate would beto require, and I repeat, require officers to breath-test all drivers involved in twocategories of events or behaviors.

The first category of events is all drivers involved in collisions. The NationalSafety Council (1995) estimated 20 million drivers were involved in collisions in 1994. Ifall collision-involved drivers were breath tested, based on Allsop’s 1966 analysis of theGrand Rapids data, I estimate apprehending an additional roughly 700,000 drivers abovethe legal limit.

A second category of behaviors that would lead to breath testing would be if adriver committed one of the 5 leading traffic infractions identified in NHTSA’s studies ofdriving behaviors associated with drinking drivers (e.g., see Harris, 1980).

Adopting laws requiring officers not to depend on their imperfect behavioralassessment of drivers, but to administer breath tests for crash involvement or trafficinfractions, will do more to increase the deterrence of law enforcement than anyimprovement, desirable as it may be, of the already efficient court conviction rate of DUIarrestees.

Legal scholars, such as Steve Simon, should surely be able to defend the requiredbreath testing for drivers committing traffic infractions or being involved in trafficcollisions as meeting the constitutional requirement for probable cause.

REFERENCES

Allsop, R. E. Alcohol and Road Accidents: A Discussion of the Grand Rapids Study. RRL Report No. 6.Road Research Laboratory, Ministry of Transport, Harmondsworth, Great Britain, 1966.

Harris, D. H. Visual Detection of Driving while Intoxicated. Human Factors, Vol. 22, No. 6, Dec. 1980,pp. 725–732.

National Safety Council. Accident Facts, 1995 Edition. Itasca, Ill., 1995.

Tashima, H. N. and C. J. Helander. 1996 Annual Report of the California DUI Management InformationSystem. Report No. CAL-DMV-RSS-96-159, Department of Motor Vehicles, Sacramento, Calif.,1996.

Tashima, H. N. and C. J. Helander. 1993 Annual Report of the California DUI Management InformationSystem. Department of Motor Vehicles, Sacramento, Calif., 1993.

Tharp, V., M. Burns, and H. Moskowitz. Development and Field Test of Psychophysical Tests for DWIArrest. Technical Report DOT HS-805-864. National Highway Traffic Safety Administration,Springfield, Va., 1981.


Urso, T., J. S. Gavaler, and D. H. Van Thiel. Blood Ethanol Levels in Sober Alcohol Users Seen in anEmergency Room. Life Sciences, Vol. 28, 1981, pp. 1053–1056.

Wells, J. K., M. A. Greene, R. D. Foss, S. A. Ferguson, and A. F. Williams. Drinking Drivers Missed atSobriety Checkpoints. Journal of Studies on Alcohol, Vol. 58, 1997, pp. 513–517.

Widmark, E. M. P. The Theoretical Foundations and the Practical Application of Forensic-MedicalAlcohol Determination. Berlin: Wien, Urban and Schwartzenberg, 1932 (English translation publishedas Principles and Applications of Medicolegal Alcohol Determination. Biomedical Publications, Davis,Calif., 1981).

Comments on “Evidence of Alcohol and DrugImpairment Obtained After Arrest”

MICHELE FIELDS


VIDEOTAPING

Videotaping driving behavior and field sobriety tests has unquestionable evidentiaryvalue for the obviously impaired driver. But it may be a trap for the unwary when used ondrivers exhibiting subtle signs of impairment. If a tape contains unambiguous evidence ofimpairment, then one would expect that it would discourage even the most aggressivedefendant from going to trial. However, if the tape does not contain evidence of obviousimpairment, at best, it will be useless to the prosecution; at worst, it will become evidencefor the defense.

In his paper, Steve Simon notes the possibility that videotape evidence can beuseful to the defense. He suggests that toxicologists should be prepared to testify thatexperienced drinkers can, and often do, show little or no outward signs of impairmenteven at high blood alcohol concentrations. This is a reasonable response to a very realproblem, but it raises its own problems. After all, the defendant is being charged with adriving offense, not with being an experienced drinker. How is the state going to showthe defendant is an experienced drinker? And if the state raises the “experienced drinker”issue, will not juries expect it to demonstrate that the defendant is such a drinker?

There are other policy concerns with regard to the use of videotape equipment. Arecent study by the National Highway Traffic Safety Administration notes that officerswith video equipment often use it selectively because they do not want ambiguousevidence for the defense to exploit (Jones et al., 1998). If videotape use becomeswidespread, then officers will be faced with two equally unacceptable alternatives. Theycan tape the subtly impaired driver, thereby collecting defense evidence, or they can useit selectively only on obviously impaired drivers. The latter course invites the defense toargue that the lack of videotape evidence suggests impairment cues were not present. Theavailability of videotape equipment could discourage officers from pursuing the driverswho do not appear to be obviously impaired.

If we significantly increase the percentage of impaired drivers who are detectedand prosecuted, then we should carefully consider the unintended consequences of theuse of videotapes, especially with regard to the subtly impaired driver.

RECOMMENDATIONS

Simon makes four excellent recommendations for improving chemical test laws:

• Amend per se laws to specify the offense is driving with a prohibited breath,blood, or urine concentration, thereby eliminating the need to express evidence frombreath or urine tests in terms of blood alcohol concentrations;

• Permit officers to choose the test to be given;

G-19


• Prohibit relation back evidence; and• Eliminate the right to independent testing while in custody.

If per se laws are constitutional, a successful challenge to the firstrecommendation is difficult to imagine. Creative lawyers could undoubtedly assertconstitutional problems with the remaining recommendations, but they would be unlikelyto succeed.

These recommendations have a common theme. They all simplify overly complexlaws. For reasons that are beyond the scope of this workshop, DUI/DWI laws havebecome exceptionally and unnecessarily complicated. Their complexity has erectedsubstantial barriers to effective enforcement.

Any measure that simplifies the burden on police in collecting and preservingevidence should be given careful consideration. Efforts that complicate the process, evenif undertaken for a legitimate purpose, should be pursued with extreme caution.

REFERENCE

Jones, R. K., J. H. Lacey, and C. H. Wiliszowski. Problems and Solutions in DWI Enforcement Systems.U.S. Department of Transportation, Washington, D.C., 1998.

Issues in the Detection of Drugs Other Than Alcohol

J. MICHAEL WALSH

The Walsh Group, P.A.

ABSTRACT

Driving under the influence of drugs is common among those arrested for DUI, and it isless frequently detected, discouraged, or treated when compared with drunk driving. Thevarious scientific and technical parameters that have restrained prevention/deterrencestrategies to deal with drugged driving are discussed in detail. To prevent drug-relatedtraffic accidents, law enforcement officials must be able to detect drivers under-the-influence of drugs as they routinely do now with alcohol detection devices. Theavailability of technologically advanced on-site drug-testing devices that are capable ofproviding immediate, accurate, and reliable results provides a unique opportunity toeffectively apply this technology in traffic safety and in the enforcement of drugs anddriving laws.

BACKGROUND

Drug abuse affects morbidity and mortality in a variety of ways. Among theunrecognized casualties are dead and injured individuals in vehicular accidents caused byor associated with operating a motor vehicle under the influence of illegal drugs.Evidence gathered over the past 50 years has established a direct relationship betweenincreasing blood alcohol concentrations (BAC) in drivers and increasing risk of motorvehicle accident (Council on Scientific Affairs, 1986). As a result, over the past 10 yearsmajor initiatives in the United States focusing on driving under the influence of alcohol(DUI) have seen a significant reduction in accidents/deaths due to alcohol intoxication(MMWR, 1997; NHTSA, 1997). To some extent this success is due to the fact thatbiochemical devices used to assess breath-alcohol-concentration are widely available, arerelatively inexpensive, and are used universally by law enforcement agencies todetermine DUI and enforce the associated laws. Unfortunately, until recently there havebeen no similar devices available to test for illegal drugs.

PREVALENCE OF DRUGGED DRIVING

In comparison with the alcohol literature, relatively little information is available regardingthe true incidence and prevalence of illegal drug use in reckless driving and drivingaccidents. Breath-alcohol testing technology has established a scientifically soundestimation of the prevalence of alcohol use among reckless drivers (Dubowski, 1992). Theprincipal problem with estimating “drugged” drivers has been the relative unavailability of

H-1

H-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

drug detection devices to test for illegal drugs. In general, such testing devices were limitedto highly specialized forensic laboratories (Turk et al., 1974).

Some data have emerged over the past 20 years, however, which give insight as tothe extent of the problem (Willette and Walsh, 1983). Lundberg et al. (1979) reported thatthe presence of psychoactive drugs other than, or in addition to, alcohol was common in apooled sample of 765 persons with driving behavior problems in Nevada and California.Williams et al. (1985) reported on a “high risk” sample of 440 young male auto driverskilled in California traffic accidents in which blood specimens collected from thesedrivers indicated 70 percent contained alcohol and more than 40 percent containedother drugs. Soderstrum et al. (1988) found that of 1,023 patients admitted to TheMaryland (Baltimore) Shock-Trauma Unit, 34.7 percent had very recently used cannabis(i.e., greater than 2ng/ml tetrahydrocannabinol in serum) and 33 percent had BAC’s greaterthan 100mg/dL. Marzuk et al. (1990) examined postmortem blood and urine of motorvehicle fatalities in New York City (1984–87) and found that at least one in four deaddrivers (ages 16–45) had used cocaine within 48 hours of death. Marzuk et al. detectedeither cocaine metabolites, alcohol, or both in 56 percent of those killed in fatal trafficaccidents in New York City.

In a collaborative effort between the National Institute on Drug Abuse (NIDA) andthe National Transportation Safety Board (NTSB) (Crouch et al., 1993), an investigation offatal-to-the-driver trucking accidents was conducted in eight states over a one-year period.Comprehensive drug screens on blood specimens collected from 168 fatally injured truckdrivers indicated that one-or-more drugs were detected in 67 percent of the drivers and33 percent of the drivers had detectable blood concentrations of psychoactive drugs oralcohol. The most prevalent drugs were cannabinoids and ethanol (each found in 13 percentof the dead drivers); cocaine was found in 8 percent of the cases, and amphetamine-likesubstances in 7 percent.

In an early evaluation of “on-site” immunoassay technology conducted inMemphis, Tennessee (Brookoff, 1994), individuals arrested for reckless driving whowere not apparently impaired by alcohol (had no odor of alcohol or tested negative bybreath analysis or both) were tested using a rapid immunoassay urinalysis test for marijuanaand cocaine at the scene of the arrest with 59 percent testing positive (18 percent for bothdrugs, 33 percent for marijuana alone, and 20 percent for cocaine alone). However, whenthe specimens were reanalyzed in a forensic laboratory, results indicated that the particularon-site device used had produced unacceptably high false positive and false negative resultsfor marijuana. Walsh et al. and Buchan et al. (1997, 1998) recently evaluated four on-sitedevices for use by police officers in testing DUI suspects. In addition to determining thefeasibility of using these devices, one of the specific aims of the project was to determinethe extent to which potentially impairing drugs are found in drivers who fail to pass aroadside sobriety test. More than 300 urine specimens were collected from DUI suspects.Each specimen was analyzed on each of four devices and then sent on to a commerciallaboratory [certified by the U.S. Department of Health and Human Services (HHS) forforensic drug testing] for reanalysis by immunoassay screen and GC/MS confirmation.The data indicated that 26 percent of the total specimens (78/303) were confirmedpositive by GC/MS in the HHS-certified laboratory for one or more illegal drugs. Of thoseindividuals who were able to pass the breathalyzer test (i.e., BrAc < 0.08), 41 percent testedpositive for one or more illegal drugs.

Walsh H-3

Studies evaluating “drugged” driving have primarily used blood or urine to makeprevalence estimates. Neither of these specimens can produce an unqualified estimate of theprevalence of “drug-impaired drivers” due to the complexities of the pharmacokinetics andpharmacodynamics of most drugs. The interpretation of drug concentrations in biologicalfluids, especially with regard to behavioral effect, requires some knowledge about the dose,the route of administration, the pattern or frequency of drug use, and the dispositionalkinetics (distribution, metabolism, and excretion) of the drug. This information is rarelyavailable in a forensic situation (Hawks and Chiang, 1987). The decline in blood/plasmaconcentrations of a drug depends on the disposition of the drug in the body, and thedisposition kinetics vary considerably across individuals. In urine the drug concentrationalso depends on dispositional kinetics and can be even more variable than that in blood orplasma because the urine volume and the urinary pH (which can effect drug elimination)may vary considerably. Interpreting the meaning of either drug/metabolite concentration ina single biological specimen (either blood or urine) with reference to impaired driverperformance is therefore an extremely difficult task. The variables involved create asufficiently great range of possible interpretations to render any specific interpretationquestionable, other than that the individual has used a specific drug in the immediatepast (days) (Hawks and Chiang, 1987). These complicated interacting pharmacokinetic/pharmacodynamic relationships have restricted the establishment of specific levels of drugconcentrations that could be interpreted as evidence of impairment either in blood, plasma,or urine (NIDA Consensus, JAMA, 1985).

DRUGS AND DRIVING BEHAVIORAL RESEARCH

A large number of laboratory and simulator studies have been conducted examining theeffects of drugs on the behavioral skills required for driving (e.g., Moskowitz, 1985;Smiley, 1981, Robbe and O’Hanlon, 1993). A detailed description of this research isbeyond the scope of this paper. In general, the behavioral research literature indicates thatdrugs can affect driving skills in many ways. The magnitude of the drug effect onbehavior is dependent on a variety of factors, including drug, dose, age, sex, weight of thesubject, prior experience with the drug, behavioral tolerance, etc. (TransportationResearch Board, National Research Council Report, 1993).

DRUG EVALUATION AND CLASSIFICATION RESEARCH

Another line of research has been the development of “Drug Evaluation andClassification” (DEC) programs designed to provide training for law enforcement officersto recognize physiological and behavioral symptomatology associated with recent druguse. The Los Angeles Police Department (LAPD) developed and implemented such aprogram, and by the early 1980s the Department was using “Drug RecognitionEvaluations” (DREs) to evaluate those suspected of driving under the influence of drugs.The DRE consists of a standardized 12-component evaluation process that providescareful observation of the suspect’s appearance, behavior, performance of psychophysicaland neurological tests, and vital signs. Those failing the DRE are required to provideblood or urine for toxicological evaluation. In 1984 the National Highway Traffic SafetyAdministration (NHTSA) and NIDA (Bigelow et al., NHTSA Report # DOT HS-806-


753, 1985) sponsored a laboratory evaluation of the DRE program at Johns HopkinsUniversity, which showed that the LAPD DREs were very accurate in identifyingsubjects under the influence of drugs and were in most cases able to identify the class ofdrug (> 91 percent). A subsequent field evaluation conducted by NHTSA (Compton, R.,NHTSA Report # DOT HS-807-012, 1986) in Los Angeles showed the DREs were 94 percent correct when they judged that a suspect had used drugs (other than alcohol)and were able to correctly identify the drug class in 87 percent of the cases. Thesesuccessful evaluations led NHTSA to develop and standardize a curriculum incollaboration with the LAPD, and in 1987 NHTSA began to sponsor the training ofpolice in other jurisdictions via a program administered by the International Associationof Chiefs of Police. The training time required for the 3-part DRE curriculum is close to80 hours, including a two-day preschool, and 56 hours of classroom work, plus on-the-job training. Because of the time and expense involved in the current training, thenationwide prevalence of certified DRE officers is limited.

In 1992, NHTSA conducted an evaluation (Preusser et al., NHTSA Report #DOTHS 808-058, 1992) of the impact of the DEC program on enforcement and adjudicationin 11 police agencies in five states. Results varied considerably across the different policeagencies. While most of the DRE drug opinions were confirmed by chemical tests andmost of the confirmed suspects were convicted, the absolute number of DRE evaluationswere significantly fewer than predicted. Across the eleven sites evaluated (AZ, CA, CO,NY, and TX) only 3–4 percent of all DWI suspects were identified as candidates for aDRE evaluation. Based on various prevalence studies, this 3–4 percent range clearlyunderrepresents the total number of drugged drivers, but since most drug use is incombination with alcohol these numbers are reasonably consistent with drug only (noalcohol) findings. Unfortunately, most state laws do not provide for additional penaltieswhen drugs are found in addition to illegal levels of alcohol. Therefore, if the suspectfails the breath-alcohol test, law enforcement officers have little reason to pursue whetherdrugs are present in the DUI suspect.

The NHTSA DRE study did find that peak activity in these drug recognitionprograms occurred in the first year or so after training, with declining activity thereafter.Various rationale were posited for this finding but nothing conclusive was determined.One could hypothesize that the lack of immediate toxicology feedback to reinforce theDRE (i.e., by having to send the specimens off to a crime lab and wait weeks for results)dampened the officers’ enthusiasm and limited the effectiveness of the drug recognitionprogram.

IDENTIFYING BEHAVIORAL CUES THAT CAN DISTINGUISH ALCOHOL FROM OTHER DRUG USE

While a significant amount of behavioral research has been conducted on the effects ofdrugs on driving skills, no specific behavioral cues have been established to distinguishdrug use from alcohol intoxication. In a recent evaluation of the DEC program designedto validate the DEC evaluation variables and to determine the accuracy of the DRE,Heishman et al. (1986) found that certain subsets of the DRC evaluation variables couldpredict the presence of specific drugs (e.g., marijuana, cocaine) more accurately than theentire DEC evaluation. However, most of these variables were physiological indices (e.g.,

Walsh H-5

nystagmus, increased blood pressure and pulse rate, pupil dilation) rather than behavioral.In a recent study of DUI suspects (Buchan et al., 1998) an attempt was made to correlatepositive drug test results with driving behaviors as listed in the arrest report. The onlycues that appeared in high frequency with positive drug tests were “weaving” and havingbeen in a “crash.”

DRUG TESTING TECHNOLOGY

Over the past 15 years workplace drug testing has expanded exponentially. In thetransportation field the U.S. Department of Transportation (DOT) has issued regulationsfor the airline, maritime, trucking, railroad, pipeline, mass transit, and other transportationindustries requiring testing for those employees in specified safety-sensitive occupations(pilots, engineers, truck drivers, etc.) These regulations cover nearly 8 milliontransportation workers. With regard to nonregulated driving, most states have the legalmeans (established in implied consent) to require drivers suspected of driving under theinfluence of drugs to submit to a drug test, but it is rarely used, primarily because crime-lab testing is typically not practical within the scope of existing workloads and priorities.The significant volume of workplace testing in the United States has created a market forsensitive, efficient, and cost-effective devices that could be used on-site for rapid analysis.

Within the past three years new technologically advanced immunoassay drug-testingdevices have been developed for use with urine specimens. These devices are capable ofrapid accurate and reliable results “on-site” and provide a unique opportunity to effectivelyapply this technology in traffic safety and in the enforcement of drugs and driving laws.

As the market for drug testing continues to expand, financial incentives aredriving the development of more sophisticated technology that is easier to use, moresensitive, and more specific. Many diagnostic manufacturers are conductingdevelopmental research on new products using alternative specimens and technologies.

A variety of specimens can be assayed for drugs (e.g., urine, blood, sweat, saliva,hair). Each specimen is unique, and each offers different patterns of information aboutdrug use over time. Figure 1 illustrates the relationship between drug effects and thedetection periods in various specimens. Each specimen has strengths and weaknessesabout the level of information that can be gained about drug use. Because the urinalysisdrug testing methodology is well established, urine has become the standard by whichother technologies are being compared. Drugs and drug metabolites are detectable inurine for several days after the drug has been used. Not only can the time window fordetection overlap with intoxication, impairment, and being “under the influence” but itcan also extend beyond these states of behavioral impairment. Therefore, while a positiveurine test is solid proof of drug use within the past few days, it cannot be used by itself toprove behavioral impairment.

Blood testing: In terms of attempting to relate drug concentrations to behavioralimpairment, blood is probably the specimen of choice. Blood has been used in someepidemiology studies; however, because of the invasiveness of the procedure and theinability of forensic toxicologists to agree on behaviorally toxic plasma concentrationslittle or no work has been done recently with blood.

Saliva testing: In comparison with urinalysis, saliva offers different informationregarding the recency of drug use. Detection times for drugs in saliva are roughly similar

H-6

TR

B C

ircular E-C

020: Issues and Methods in the D

etection of Alcohol and O

ther Drugs

FIGURE 1 Drug Effects and Detection Periods (From E. J. Cone, Addiction ResearchCenter, NIDA)

Walsh H-7

to that in blood (approximately 4–24 hours). Collection of saliva is generally consideredless invasive than either blood or urine and can be an excellent matrix to tie recent druguse with behavioral impairment. The negative side of saliva testing is that testing must bedone in a laboratory, specimen collection can be easily contaminated, and the methodsfor assaying saliva for most drugs of abuse are still in the developmental stages (Acocaine assay has recently been approved by the Food and Drug Administration (FDA),and Opiate and PCP assays are under review; however, no assay for THC is available.).These assays are being developed by STC Technologies (Bethlehem, PA) for use with theOraSure Oral Collection Device manufactured by Epitope, Inc. (Beaverton, OR). At thistime, no nationally established standard methods for saliva drug testing or anycertification programs are currently available. However, rapid on-site saliva tests foralcohol are available and have been approved by the FDA. Some on-site alcohol deviceshave been included by NHTSA on their conforming products listing as suitable for use asscreening test devices in the DOT workplace testing programs.

Sweat testing: Sweat-testing methods have recently been approved by the FDA,which include a sweat patch collection device. This patch is designed to collect drugs ofabuse from human skin. The patch can be worn for periods up to several weeks, followedby removal and laboratory analysis. This device can measure cumulative drug use overtime. The manufacturer (PharmChem Labs, Menlo Park, CA) is developing a patch forworkplace usage, which could generate a sweat sample in a 20-minute period. Such adevice could provide for the detection of recent drug use, although the detection timewindow for drugs in sweat has not yet been clearly established. The cons of sweat testinginclude the high variability of results among individuals, due to the low concentrations ofanalytes detectable in sweat.

Hair testing: While the technology for assaying hair for drugs of abuse hasprogressed somewhat over the past 15 years, there remain many unresolved issues (e.g., itremains unclear how drugs actually enter the hair). Because hair only grows at a rate ofabout one-half inch per month, it is not suitable for the detection of recent use. Therefore,it is highly unlikely that hair could serve as a viable specimen in DUI-drug testing.

Urine testing: At this writing it appears that the most viable method for accurate andreliable drug testing is urinalysis. Laboratory-based urine testing for drug of abuse is widelyavailable, and HHS certifies laboratories for forensic drug testing. The DOT regulationsrequire the use of these federally certified labs. Currently, HHS has nearly 80 labs in theprogram that are cumulatively processing about 70,000 specimens per day at an averagecost in the range of $8–20, depending on the volume.

On-site urine testing: Some of the most recent advances in drug testing have beenthe developments in the on-site testing products. There are at least 17 rapid on-siteimmunoassay devices currently available on the commercial market. These devices havebeen designed primarily for workplace testing. In a recent evaluation (Walsh et al., 1997)four of these devices were evaluated for use by police officers in testing DUI suspects.Each of the on-site immunoassay devices worked well, although there were differences inthe ease of handling and complexity of the test. The authors concluded that several ofthese devices were quite suitable for use as screening devices and that law enforcementofficers were capable of learning to use these devices in a very skilled manner whenproperly trained.


ISSUES IN THE DETERMINATION OF “DRIVER IMPAIRMENT” DUE TODRUGS AND “DRIVING UNDER THE INFLUENCE OF DRUGS”

I believe that with the prevalence of illegal drug use in the nation, driving under theinfluence of drugs in addition to alcohol has become a significant public health problem.As noted by others (Saylor et al., 1992), driving under the influence of drugs is commonamong those arrested for DUI and is less frequently detected, discouraged, or treatedwhen compared with drunk driving. In fact, most state laws do not provide for additionalpenalties when drugs are found in addition to illegal levels of alcohol. Therefore, if thesuspect fails the breath-alcohol test, law enforcement officers have little reason to pursuewhether drugs are present in the DUI suspect. As a result of the growing body ofevidence of illegal drug use by drivers, a number of states (AZ, GA, IL, MN, RI, UT andIN) have enacted per se laws that generally imply that drug concentrations in body fluidscan be used as evidence of being “under the influence” of a drug (Lewis and Buchan,1998). Under such statutes individuals can be found guilty of “driving under theinfluence” if he or she was operating a motor vehicle while any illicit drugs were presentin his or her system. The concept of “driving under the influence” creates an importantlegal distinction between proving that observed driver impairment is due to taking a drug(causal relationship) and proving that observed impaired driving behavior was associatedwith specified concentrations of drug/metabolite in the individual’s body fluids.

For more than 20 years the scientific community has been aware of this growingproblem of drugged driving but has been reluctant to make strong recommendationsbecause of the complexities of assaying drugs in body fluids and interpreting the results.The 1985 NIDA consensus report (JAMA, 1985) concluded that little is known and moreresearch is needed. The NIDA report did lay out a blueprint for research that could leadto the development of prevention/intervention strategies:

In order to establish that use of a drug results in impairment of driving skills andto justify a testing program to respond to this hazard, certain facts must be avail-able. 1. The drug can be demonstrated in laboratory studies to produce a dose-related impairment of skills associated either with driving or with relatedpsychomotor functions. 2. Concentrations of the drug and/or its metabolites inbody fluids can be accurately and quantitatively measured and related to thedegree of impairment produced. 3. Such impairment is confirmed by actual high-way experience. 4. Simple behavioral tests, such as can be done at the roadside bypolice officer with modest training, can indicate the presence of such impairmentto the satisfaction of the courts. 5. A range of concentrations of the drug can beincorporated in laws relating to impaired driving as ipso facto evidence.

Nearly 15 years have passed since the NIDA conference and while we now havemuch of the recommended information, we still do not have all the answers. Theproblems of individual differences and the variability in the pharmacokinetics /pharmacodynamics of most drugs make it virtually impossible to tie drug specific drugconcentrations to behavioral impairment.

I do not believe that this gap in our knowledge base is sufficient reason tocontinue to ignore the problem. To prevent drug-related traffic accidents, lawenforcement officials must be able to detect drivers under the influence of drugs as they

Walsh H-9

routinely do now with alcohol detection devices. The availability of technologicallyadvanced on-site drug-testing devices that are capable of providing immediate accurateand reliable results provides a unique opportunity to effectively apply this technology intraffic safety and in the enforcement of drugs and driving laws. The routine availability ofdrug testing for DUI and of DRE officers could be used as powerful prevention tools todetect and deter drugged driving behavior.

REFERENCES

Brookoff, D., C. Cook, C. Williams, and C. Mann. Testing Reckless Drivers for Cocaine and Marijuana.New England Journal of Medicine, Vol. 331, 1994, pp. 518–522.

Buchan, B. J., J. M. Walsh, and P. E. Leaverton. Evaluation of the Accuracy of On-Site Multianalyte DrugTesting Devices in the Determination of the Prevalence of Illicit Drugs in Drivers. Journal of ForensicSciences, Vol. 43, No. 2, 1998, pp. 395–399.

Consensus Development Panel. Drug Concentrations and Driving Impairment. Journal of the AmericanMedical Association, Vol. 254, No. 18, 1985, pp. 2618–2621.

Council on Scientific Affairs. Alcohol and the Driver. Journal of the American Medical Association, Vol. 255,1986, pp. 522–527.

Crouch, D., M. Birkey, S. Gust, D. Rollins, J. M. Walsh, J. Moulden, K. Quinlan, and R. Beckel. ThePrevalence of Drugs and Alcohol in Fatally Injured Drivers. Journal of Forensic Sciences, Vol. 38, No. 6, 1993, pp. 1342–1353.

Dubowski, K. M. The Technology of Breath-Alcohol Analysis. DHHS Publication No. ADM-92-1728.U.S. Department of Health and Human Services, Washington, D.C., 1992.

Hawks, R. L., and C. N. Chiang. Urine Testing for Drugs of Abuse. NIDA Research Monograph No. 73,DHHS Publication No. ADM 87-1481. U.S. Department of Health and Human Services, Washington,D.C., 1987.

Heishman, S. J., E. G. Singleton, and D. J. Crouch. Laboratory Validation Study of Drug Evaluation andClassification Program: Ethanol, Cocaine, and Marijuana. Journal of Analytical Toxicology, Vol. 20,Oct. 1986, 468–483.

Lewis, M. F., and B. J. Buchan. The Drugged Driver and the Need for a per se Law. Florida Bar Journal,July/Aug., 1998.

Lundberg, G. D., J. M. White, and K. I. Hoffman. Drugs (other than or in addition to Ethyl Alcohol) and Driving Behavior: A Collaborative Study of the California Association of Toxicologists. Journalof Forensic Sciences, Vol. 24, 1979, pp. 207–215.

Marzuk, P. M., K. Tardiff, A. C. Leon, M. Stajic, E. B. Morgan, and J. J. Mann. Prevalence of RecentCocaine Use among Motor Vehicle Fatalities in New York City. Journal of the American MedicalAssociation, Vol. 236, No. 2, Jan. 12, 1990.

Morbidity and Mortality Weekly Report. Vol. 46, No. 48, Dec. 5, 1997.

Moskowitz, H. A. Marijuana and Driving. Accident Analysis and Prevention, Vol. 17, 1985, pp. 323–346.

National Highway Traffic Safety Administration. Traffic Safety Facts, 1996. NHTSA, U.S. Department ofTransportation, 1997.

Robbe, H. W. J., and J. F. O’Hanlon. Marijuana and Actual Driving Performance. Report No. 808078.NHTSA, U.S. Department of Transportation, 1993.

Saylor, K. E., R. L. Dupont, and H. Brown. The High Way: Driving under Influences Other Than Alcohol.Journal of the American Medical Association, Vol. 267, 1992, p. 652.


Smiley, A. M. Marijuana: On-Road and Driving Simulator Studies. Alcohol, Drugs, and Driving, Abstractsand Reviews, Vol. 2, 1981, pp. 121–134.

Smiley, A. M., H. A. Moskowitz, and K. Ziedman. Driving Simulator Studies of Marijuana Alone and inCombination with Alcohol. In Proceedings of the 25th Conference of the American Association ofAutomotive Medicine, 1981, pp. 107–166.

Soderstrum, C., A. Trifillis, B. Shankar, W. Clark, and R. Cowley. Marijuana and Alcohol Use Among1023 Patients. Archives of Surgery, Vol. 123, June 1988, pp. 733–737.

Turk, R. F., A. J. McBay, and P. Hudson. Drug Involvement in Automobile Driver and PedestrianFatalities. Journal of Forensic Sciences, Vol. 19, No. 1, Jan. 1974.

Walsh, J. M., B. J. Buchan, and P. E. Leaverton. Detection of Illicit Drugs in Drivers. In Proceedings of the14th International Conference on Alcohol, Drugs and Traffic Safety, Vol. 2, 1997, CERMT, Annecy,France, pp. 485–491.

Willette, R. E., and J. M. Walsh. Drugs, Driving and Traffic Safety. WHO Publication No. 78. WorldHealth Organization, Geneva, Switzerland, 1983.

Williams, A. F., M. A. Peat, D. J. Crouch, J. K. Wells, and B. S. Finkle. Drugs in Fatally Injured YoungMale Drivers. Public Health Reports, Vol. 100, No. 1, 1985, pp. 19–25.

Comments on “Issues in the Detection of Drugs Other Than Alcohol”

HAN DE GIER

University of Utrecht

H-11

In reflecting on the paper by Dr. Walsh and in addressing the issues of this workshop, Iwould like to focus on drugs other than alcohol. With respect to illicit drugs I would liketo discuss in brief some recent developments in Europe (particularly in Belgium), wherenew legislation has been proposed to prosecute drivers with measurable amounts of illicitdrugs and/or metabolites in their system. With respect to licit drugs, it is clear that theyhave been excluded in most discussions so far, and also in Dr. Walsh’s paper. This isprobably because no one actually knows what to do with them, especially in establishinga basis for criminal prosecution. I will comment a little more on prevention in the case ofmedicinal drugs, although this is somewhat beyond the scope of this workshop.

Dr. Walsh has presented a clear overview of problems that will be encountered indiscussing issues related to illicit drugs. He does not cover licit drugs to the extent that wecan discuss them with background information similar to that he has given on illicit drugs.Some of Dr. Walsh’s conclusions are the same for both categories of drugs, such as

1. The interpretation of the meaning of either drug and/or metabolite concentrationsin a single biological specimen (either blood or urine) with reference to impaired driverperformance.

2. The variety of factors that determine the magnitude of the drug’s effect onbehavior (including dose, sex, body weight, prior experience, development of tolerance).

3. The lack of provisions for additional penalties in most laws when drugs are foundin addition to alcohol, and the impact this has on the efforts of law enforcement officersto look for drugs other than alcohol in cases where alcohol was found positive.

4. The fact that recent technological developments provide us with advanced drugtesting devices for use “on-site,” in particular for drug screening in urine. Dr. Walshpresented only a few devices that include licit drug testing, such as for benzodiazepinesand tricyclic antidepressants.

In general, we have limited knowledge on the prevalence of drugs other than alcoholin road traffic. Most reports deal with small sample sizes taken from selected driverpopulations and do not reflect the general driver population. Without discussing themethodological issues dealing with the selection of driver populations (general, suspectedof DUI of drugs, collision-involved), one can estimate that the prevalence of illicit druguse in the general driver population will fall (at least in Europe) in the range of 1–5percent, whereas the prevalence of licit drugs affecting driving performance will behigher (5–15 percent). The illicit drugs of interest at this moment in Europe seem to becannabis and opiates (not in particular cocaine, found in many studies in the UnitedStates), followed by amphetamines. For the licit drugs, benzodiazepines arepredominantly found, whereas tricyclic antidepressants are much-less detected. Lookingat the data for populations of drivers suspected of driving under the influence of drugs


and for collision-involved drivers, there are no clear differences reported. However,larger variations exist (10–70 percent) depending upon the selection of drivers includedin the samples. For example, in Norway the police force seems to be focused very muchon drugs other than alcohol, which causes large differences in prevalences of drug useamong drivers in comparing the results from various Nordic countries (De Gier, 1998).

The use of the combination of drugs and alcohol in the general driver populationrevealed major differences while looking at licit and illicit drug use in one large-scaleGerman roadside survey (Krüger et al., 1995). The prevalence of the combination of licitdrugs and alcohol was extremely low (only one case), whereas high prevalence wasdetected for the combination with illicit drugs (44 percent). Similarly, a high prevalenceof combined use was found in other driver populations (drivers suspected of DUI ofdrugs and collision-involved drivers) in a few other large-scale European studies.Although these studies do not all separate licit and illicit drug use, one has to concludethat the combination of drugs with alcohol is one of great concern in terms of trafficsafety. This topic has not been discussed by Dr. Walsh. The importance of the synergisticinteraction of alcohol and drugs has been stressed by several experts, both in the fields ofepidemiology and experimental human psychopharmacology. If mortality was taken asthe outcome variable, then Belgian researchers recently indicated a relative risk of 3.56 inthe combined positive group, in which a mere additive effect would theoretically have ledto a relative risk of 1.60 (Meulmans et al., 1997). Furthermore, it has been suggested thatalcohol and cannabis use in combination carries a greater risk potential than either ofthem alone, in both epidemiological research (Terhune et al., 1992) and experimentalresearch (Robbe and O’Hanlon, 1998).

These conclusions allow me to address the first of my three major comments forfurther discussion. If we know this greater risk potential on one hand and we still haveproblems deciding on drug levels for prosecution on the other, then why do we notconsider a lower legal BAC level (e.g., 0.02 percent instead of 0.05 percent, which is thelegal limit in most European countries for those drivers who are screened positive foralcohol and drugs)? If we can extrapolate the conclusions of the German roadside survey,this will primarily affect the drivers using illicit drugs, since none of the drivers who werefound positive for benzodiazepines in the German study showed positive screeningresults for alcohol as well. Maybe patients using these drugs are more aware of thepotential risks of using their medication together with alcohol. The approach to loweringthe legal limit might also contribute in part to eliminating the concern of Dr. Walsh. Inhis paper he indicated clearly that the lack of immediate penalties when drugs are foundwith low blood alcohol levels will limit the effectiveness of law enforcement anddiminish the enthusiasm of officers who are aware of problems with drugs other thanalcohol and driving.

My second comment for discussion is the redefinition of the concept of “drivingunder the influence.” The road traffic laws in Belgium and Germany will be changed soon:not impairment, but the presence of the drug in the system has to be proved for prosecution.The two countries have in fact decided in favor of the zero tolerance option, but only forillicit drugs. Experts did not have to argue about “impairment levels” but just to decide onspecified concentrations regarding analytical “cut-off” values. The detection process startswith the screening of urine samples for drugs, followed by confirmation in blood tests.

Gier H-13

My third and last comment for discussion concerns licit or medicinal drugs andthe concept of “illegal drug use,” a terminology frequently used by Dr. Walsh in hispaper. If a drug is prescribed by a physician in treating a medical condition, how can wespeak about illegal drug use? The patient’s use of a drug can only be considered “illegal”in cases in which the prescribing physician has forbidden the patient to drive, forexample, during a certain period of time. The rationale for this advice to the patient isgiven by Neutel (1995), from Canada. He performed a pharmacoepidemiological studyand revealed an extremely high risk of accident involvement the first two weeks (morethan a 10-fold increase during the first week, and a 5- to-6-fold increase the second week)of using a first prescription for a benzodiazepine tranquilizer or hypnotic. If we couldenforce prescribing physicians and dispensing pharmacists to act in accordance withknowledge from Neutel’s study, then we could decrease the number of traffic injuries andfatalities. In addition, more prevention is possible if we can change their practices towardthe use of safer or the least-impairing alternatives (which do exist in the major therapeuticclasses). These preventive measures could better serve the search for solutions thandiscussions of how to tie specific drug concentrations to behavioral impairment as part ofthe detection process. The research needs in the area of pharmacoepidemiology are fairlysimple to fulfill in those countries where drug exposure data and accident data can belinked. The results will convince the public and the policy makers that prevention andbetter pharmaceutical care are feasible options in addressing the solutions for the drug-and-driving problem. Interventions at the start of a drug treatment will probably be moreeffective than emphasizing the prosecution issues after impaired driving is detected,caused by medication prescribed without legal provisions.

REFERENCES

De Gier, J. J. Review of Investigations of Prevalence of Illicit Drugs in Road Traffic in Different EuropeanCountries. DGC, Oosterhout, The Netherlands, 1998.

Krüger, H.-P., E. Schulz, and H. Magerl. The German Roadside Survey 1992–1994. Saliva Analyses froman Unselected Driver Population: Licit and Illicit Drugs. In Proceedings of the 13th InternationalConference on Alcohol, Drugs and Traffic Safety, 1995, pp. 55–62. NHMRC, Road Accident Unit,University of Adelaide, Australia.

Meulmans, A., P. Hooft, L. Van Camp, N. De Vrieze, W. Buylaert, A. Verstraete, and M. Vansnick.Belgian Toxicology and Trauma Study. BeEDim/BIVV/IBSR/BLT, Belgium, 1997.

Neutel, C. I. Risk of Traffic Accident Injury after a Prescription for a Benzodiazepine. Annals of Epidemiology, Vol. 5, 1995, pp. 239–244.

Robbe, H. W. J., and J. F. O’Hanlon. Marijuana, Alcohol and Actual Driving Performance. NHTSA, U.S.Department of Transportation, 1998.

Terhune, K. W., C. A. Ippolito, D. L. Hendricks, J. G. Michalovic, S. C. Boegema, P. Santinga, R. Blomberg, and D. F. Preusser. The Incidence and Role of Drugs in Fatally Injured Drivers. ReportDOT HS-808-065. NHTSA, U.S. Department of Transportation, Oct. 1992.

Comments on “Issues in the Detection of Drugs Other Than Alcohol”

RICHARD P. COMPTON

National Highway Traffic Safety Administration

H-14

Mike Walsh has just provided a fairly comprehensive overview of the issues involved indetecting drug-impaired drivers, acquiring evidence that a driver is impaired or under theinfluence of drugs other than alcohol, and collecting evidence for the successfulprosecution of drug-impaired drivers. I would like to discuss briefly what I see as theimmediate research needs in the area of detecting and sanctioning the drug-impaireddriver and present some preliminary results of a survey we did on drug-impaired drivingin the United States.

DETECTING DRUG-IMPAIRED DRIVERS

There is a clear need for better methods of detecting drug-impaired drivers. While thenature and extent of the highway safety problem due to drugs other than alcohol are lesswell documented than we might like, there is ample evidence that it is a serious problem onour roads. Survey (self-report) data, incidence studies, trauma center data, and other studiesall indicate that many drivers drive while under the influence of drugs. These drivers areoften impaired, and as a result they get into crashes. Yet, there are virtually no scientificallysound tools available to law enforcement to help them detect the drug-impaired driver.

I think it is obvious to everyone that the current approach used to detect the drug-impaired driver can be fairly characterized as the “not-alcohol approach,” By this I meanthat there is no systematic approach to detecting the drug-impaired driver other than bydefault when a law enforcement officer has eliminated alcohol as a probable cause of adriver’s impairment. When an officer begins to develop a suspicion that a motorist isimpaired, as the result of observed driving behavior, behavior after the motorist has beenstopped, or the result of a roadside sobriety test, the natural inclination is to think theimpairment is due to alcohol. Only when additional evidence is accumulated, typically asa result of a breath-alcohol test that is incompatible with the observed impairment, doesthe officer realize the impairment is due to some other reason. At this point attention isturned to drugs other than alcohol.

It appears that there has been enough information accumulated through researchover the last decade to make it feasible to develop driving and behavioral cues fordetecting driving impairment due to drugs other than alcohol. This information needs tobe carefully reviewed with an eye to developing practical tools, similar to those that havebeen developed and validated for alcohol, that assist law enforcement officers to detectimpairment by the other major categories of drugs.

LEGAL ISSUES IN PROSECUTING DRUG-IMPAIRED DRIVERS

It is clear that we are unlikely to be able to deal with the drug-impaired driving problemin the same fashion we have taken with alcohol. For all the reasons Mike mentioned in

Compton H-15

his paper, it is unlikely we will have an analog to a simple BAC test, with knownconcentrations of a metabolite clearly related to driving impairment any time soon, ifever, for most drugs.

Thus, we need to recognize that a different approach is necessary to deal with thedrug-impaired driver. I wish I could lay out for you today in full detail what this approachshould be, but I cannot. However, I think the outline of what form their approach willtake, in part, is becoming apparent. There will be a clear difference in the way legal andillegal drugs are treated. Driving while under the influence of illegal drugs can beprohibited, and in fact has been in a number of states. These laws do not referenceimpairment. Evidence is still needed that the driver was under the influence of the drug,but there is no need to relate drug concentration to impairment. More widespreadadoption of these “under the influence” laws is needed. They greatly simplify the job ofthe police officer, toxicologist, and prosecutor.

This approach will not work for the legal, yet potentially impairing drugs. Thisincludes many sedatives, depressants, and stimulants, among other prescription and non-prescription drugs. Drivers showing obvious signs of impairment that is not clearly due toalcohol or illegal substances, who are under the influence of drugs, should be in violationof the law. This is an area that needs work.

Other issues that need to be addressed include the elimination of one-test laws thatmany states currently have in their statutes. These laws, under implied consent provisions,limit the state to requesting a single test of a body fluid. Alcohol is by far the single mostprevalent drug impairing motorists. It is logical for a police officer to conduct an alcoholtest first; it is most likely to be positive and is easier and cheaper to conduct. However,when the alcohol test result is negative or incompatible with the observed impairment, theoption of testing for other impairing substances should be available. The recentdevelopment of onsite screening tests using urine offers considerable promise in providingthe police with fairly quick and reliable evidence of drug use. The potential benefits ofmore widespread use of these on-site screening tests need to be looked at carefully.

PRELIMINARY HIGHLIGHTS FROM THE 1996 NATIONAL HOUSEHOLDSURVEY ON DRUG ABUSE

I would now like to share with you some preliminary results from a nationallyrepresentative survey that investigated the prevalence and patterns of drug use amongdrivers in the United States. This information concerns illicit drug use and driving and isderived from data collected in the 1996 National Household Survey on Drug Abuse’s“Driving Behaviors Module.” The driving behaviors module was developed andincorporated into the 1996 National Household Survey on Drug Abuse through acollaborative effort between the Substance Abuse and Mental Health ServicesAdministration (SAMHSA) and the National Highway Traffic Safety Administration.

The driving behavior module was designed to measure the frequency of drivingwithin two hours of illicit drug and /or alcohol use, and the circumstances of these events.It covered the use of only certain, highly prevalent illegal drugs, including marijuana,cocaine, tranquilizers, sedatives, and stimulants. The respondents were 11,847 persons,age 16 or older, who reported driving within the past 12 months. These respondentsrepresent more than 166 million drivers in the United States.


The results of this survey showed that approximately 5 percent (representingapproximately 8.9 million drivers) reported driving within two hours of illicit drug use inthe past year (see Figure 1). Twenty-three percent (representing approximately 38 milliondrivers) reported driving after alcohol use, but never after using drugs other than alcohol.Drivers who drove within 2 hours of illicit drug use were younger than drivers who didnot drive within two hours of using alcohol or drugs (see Figure 2); were more likely tobe male (see Figure 3); more likely to be single (see Figure 4); and more likely to beunemployed. When compared with drivers who drove within two hours of alcohol useonly (and the overall driving population), a higher percentage of those reported drivingafter illicit drug use were young, single, had an annual income less than $10,000, weremale (compared with all drivers), and were unemployed (compared with all drivers). Thefull results of this survey should be released sometime this fall.

FIGURE 1 Distribution of drivers ages 16and older by whether they drove within twohours of substance use in past year. Source:SAMHSA, NHSDA, 1996

FIGURE 2 Distribution of drivers ages 16and older who drove within two hours ofsubstance use by age and substance used.Source: SAMHSA, NHSDA, 1996

Compton H-17

FIGURE 3 Percentage of drivers ages 16and older who drove within two hours ofsubstance use by gender. Source: SAMHSA, NHSDA, 1996.

FIGURE 4 Percentage of drivers ages 16and older who drove within two hours ofsubstance use by marital status. Source: SAMHSA, NHSDA, 1996.

Prosecuting and Adjudicating DWI Detection Evidence

JOEL A. WATNE

State of Minnesota

ABSTRACT

Defendants and their attorneys challenge criminal and administrative sanctions byasserting noncompliance with each requirement of statute or rule and by violating stateand federal constitutional provisions. Peace officers need thorough instruction in theapplicable law governing apprehension, investigation, arrest, and testing, as well astraining to detect impaired drivers, preserve evidence, record observations, andprepare for trial. Among disincentives to increased enforcement efforts are lack of publicsupport, lack of support from the leadership in some communities, and officers’perceptions that laws are too complicated to understand and administer and thatprocedures are unduly time-consuming. Enforcement efforts and results can be enhancedby better use of existing technology and simplification of statutory provisions andprocedures, along with continued efforts to educate the public to be less tolerant ofimpaired drivers and more willing to intervene to prevent driving by the impaired or tocall the police when an impaired person insists on driving.

HOW IS DETECTION EVIDENCE CHALLENGED?

Introduction

After officers have completed the detection and investigatory stages of a DWI case,1

the case moves to the next stage: administrative or judicial hearings in administrativedriver’s license revocation2 and criminal proceedings. Here the typical challenge isbased upon a claim that the evidence sought to be used was obtained “illegally,”whether in violation of state statute or state and federal constitutional requirements.As states have increased the sanctions for DWI offenses, there has been an observableincrease in the willingness of drivers to devote more resources to contesting thosesanctions, and an increase in the number of attorneys and forensic experts offeringtheir services to those drivers.

I-1

1 Because the term “DWI” is well understood nationally and internationally, that abbreviation will be used hereininstead of the various abbreviations used in a number of states to track the precise language of their statutes such asOWI, OMVUI, DUIL, DUI, UBAL, etc. However, there are some who consider “DWI” to be legally insufficient togive notice to a person as to what offense is charged. See, e.g., State v. Raley, 86 N.M. 190, 192, 521 P.2d 1031, 1033(N.M. App. 1974) (the only known abbreviation of “D.W.I.” legally is “died without issue”).2 While some distinguish between implied consent driver’s license revocations for test refusals and “administrativedriver’s license revocations” for those who submit to testing that discloses prohibited levels of alcohol and other drugs,both groups of cases are lumped together for purposes of this paper.

I-2 TRB Circular E-C020: Issues and Methods in the Detection of Alcohol and Other Drugs

Statutory Challenges

Because the statutes of the 50 states vary in so many specific details, no attempt ismade here to catalog the possible statutory challenges under the laws of each state.Because of the explosion in appellate litigation of DWI-related issues in recentdecades, references here are largely limited to Minnesota cases. One commonrequirement in many states in the past has been that an arrest could be made only ifthe offense was committed in the officer’s “presence.”3 In Minnesota, for example,there have been challenges based on proof that the officer who requested a test under the implied consent law was a “peace officer” as defined in the impliedconsent law;4 on whether a bulldozer being operated on private property was a “motorvehicle” to which the implied consent law applied;5 on whether the “physical control”provisions applied to the conduct involved;6 on whether the Implied Consent Advisoryform must be read in the squad car at the scene of arrest or at the location where thetest will be administered;7 on whether a Minnesota officer may follow a driver injuredin a Minnesota accident to a Fargo, North Dakota hospital, invoke the Minnesotaimplied consent law in North Dakota, and obtain a test in North Dakota;8 on whetheran interpreter must be provided for a driver who speaks only Japanese or is deaf;9 onwhether the officer offered the driver a choice of testing as prescribed in the statute;10

on whether the driver has the right to demand a blood test instead of a breath test;11

3 See, e.g., State v. Cormican, 292 Minn. 505, 195 N.W.2d 586 (1972). In Minnesota, the “misdemeanor presence rule”was repealed in 1984, and arrests have been authorized on “probable cause” alone since then.4 See State, Dept. of Highways v. Halvorson, 288 Minn. 424, 181 N.W.2d 473 (1970).5 See Ruzic v. Commissioner of Public Safety, 455 N.W.2d 89 (Minn. Ct. App. 1990) (question answered in the affirmative).6 See, e.g., State, Dept. of Public Safety v. Juncewski, 308 N.W.2d 316 (Minn. 1981) (legislature intended statute to be givenbroadest possible application and effect); State v. Woodward, 408 N.W.2d 927 (Minn. Ct. App. 1987) (driver standing byrear bumper of her car, which had a flat tire; claimed someone else drove it there); Abeln v. Commissioner of Public Safety,413 N.W.2d 546 (Minn. Ct. App. 1987) (driver asleep in driver’s seat of vehicle, with dead battery, was in physicalcontrol); Short v. Commissioner of Public Safety, 422 N.W.2d 40 (Minn. Ct. App. 1988) (person asleep in driver’s seatof bar parking lot, key in ignition, claiming he entered vehicle solely to use car phone, was in physical control); Berns v.Commissioner of Public Safety, 355 N.W.2d 493 (Minn. Ct. App. 1984) (driver in driver’s seat, removed keys and put themon floor by passenger’s feet to avoid being in physical control, was in physical control); Dufrane v. Commissioner of PublicSafety, 353 N.W.2d 705 (Minn. Ct. App. 1984) (driver asleep in driver’s seat, key never found, was in physical control).7 See Rohlik v. Commissioner of Public Safety, 400 N.W.2d 791 (Minn. Ct. App. 1987) (rejecting claim that officer wasrequired to read the form in the squad car at the scene of arrest rather than at the hospital); Bird v. Commissioner ofPublic Safety, No. C7-97-1090 (Minn. Ct. App. Feb. 3, 1998) (unpublished) (rejecting claim that it was improper toread the advisory in the squad car at the scene of the arrest).8 See, e.g., State v. Torgerson, 453 N.W.2d 698 (Minn. 1990) (question answered in affirmative); Boland v.Commissioner of Public Safety, 520 N.W.2d 487 (Minn. Ct. App. 1994) (same).9 See, e.g., Yokoyama v. Commissioner of Public Safety, 356 N.W.2d 830 (Minn. Ct. App. 1984) (need not provideinterpreter or advisory warnings in Japanese); Warner v. Commissioner of Public Safety, 498 N.W.2d 285 (Minn. Ct.App. 1993) (making interpreter available for deaf driver is desirable, but not essential and should not interfere withevidence gathering).10 See, e.g., State, Department of Highways v. McWhite, 286 N.W.2d 468, 176 N.W.2d 285 (1970) (must offer bloodtest); Haugen v. Commissioner of Public Safety, 389 N.W.2d 222 (Minn. Ct. App. 1985) (must offer choice betweenblood and urine tests if either is offered); Workman v. Commissioner of Public Safety, 477 N.W.2d 539 (Minn. Ct. App.1991) (need not offer choice of alternative test unless first offer is refused). See also Moe v. Commissioner of PublicSafety, 574 N.W.2d 96 (Minn. Ct. App. 1998) (“due process” does not require that driver be given a choice of tests).11 See, e.g., Forrest v. Commissioner of Public Safety, 366 N.W.2d 371 (Minn. Ct. App. 1985); Carlson v.Commissioner of Public Safety, 357 N.W.2d 391 (Minn. Ct. App. 1984), both answering the question in the negative.

Watne I-3

and on whether the officer violated the statutory right to an additional or independenttest,12 among others.

The most obvious conclusion from reading the cases in Minnesota and other statesis that the more requirements the legislature imposes in a statute, the greater the likelihoodof an error that provides an opportunity for challenge. The most obvious remedies are (1) tosimplify statutory requirements and eliminate those that are unnecessary orcounterproductive, and (2) to provide additional training to officers so that they fullyunderstand the requirements they must meet.13

Challenges Under State Constitutions

In recent years, appellate courts in a number of states have chosen to exercise theirauthority to provide greater protection to DWI suspects under the state constitution thanis available under the Constitution of the United States of America. The motivations caninvolve either a fundamental ideological disagreement with the decisions of the U.S.Supreme Court or simply the personal views of individual members of an appellate court.The principal advantage of resorting to the state constitution is that a state supreme courtis the final authority in its interpretation, effectively insulating the decision from anappeal to the U.S. Supreme Court.

Thus, while the U.S. Supreme Court held that states are free to use sobrietycheckpoints and that the court will not second-guess the decisions of state authorities asto the relative effectiveness of various strategies in detecting DWI offenders,14 theMinnesota Supreme Court imposed a requirement that the state prove the checkpoints tobe more effective than traditional roving patrols and held the checkpoint in question to bein violation of the state constitution.15

Although DWI suspects have no federal constitutional right to consult with anattorney before deciding whether to submit to testing under the implied consent statute,16

the Minnesota Supreme Court recognized such a right under the state constitution.17

12 See, e.g., State v. Streitz, 276 Minn. 242, 150 N.W.2d 33 (1967) (no duty to place call for driver); Frost v.Commissioner of Public Safety, 348 N.W.2d 803 (Minn. Ct. App. 1984) (officer need not talk to doctor regardingarrangements for test, or keep driver in custody until doctor arrives); Hager v. Commissioner of Public Safety, 382N.W.2d 907 (Minn. Ct. App. 1986) (officer need not transport driver to hospital for test); Theel v. Commissioner ofPublic Safety, 447 N.W.2d 472 (Minn. Ct. App. 1989) (officer must not discourage driver from attempting to arrange foradditional test).13 In May 1998, this writer and fellow presenter Steven Simon attended a meeting in Minneapolis between statetroopers, deputy sheriffs from several counties, police officers from numerous jurisdictions, prosecutors and legislators,organized by MADD, in which the loudest complaint of the officers was that the Minnesota laws had become socomplex and involved so much paperwork and court time that it was very difficult for officers to comply with everyrequirement no matter how hard they tried. Some officers have expressed so much frustration with following suchcomplex requirements that they, or their colleagues, have given up actively looking for DWI violators and are dealingonly with those too obvious to ignore.14 See Michigan Department of State Police v. Sitz, 496 U.S. 444, 110 S.Ct. 2481 (1990).15 See Ascher v. Commissioner of Public Safety, 519 N.W.2d 183 (Minn. 1994). However, in Ascher v. Commissionerof Public Safety, 527 N.W.2d 122 (Minn. Ct. App. 1995), review denied (Minn. Mar. 21, 1995) (“Ascher II”), the courtheld that the evidence obtained in that unconstitutional sobriety checkpoint could be used to cancel the same driver’slicense for violating the total abstinence restriction under which he had previously regained driving privileges.16 See Nyflot v. Commissioner of Public Safety, 369 N.W.2d 512 (Minn. 1985); Prideaux v. State, Dept. of PublicSafety, 310 Minn. 405, 247 N.W.2d 385 (1976); State v. Palmer, 291 Minn. 302, 191 N.W.2d 188 (1971).17 See Friedman v. Commissioner of Public Safety, 473 N.W.2d 828 (Minn. 1991).


Also, while the U.S. Supreme Court has held that no Fourth Amendment“seizure” has occurred when a suspect is fleeing officers until the person either is touchedby the pursuing officers or has submitted to their show of authority,18 the MinnesotaSupreme Court has held that merely commanding the suspect to stop is sufficient toconstitute a “seizure.”19

When such a decision adversely affects DWI enforcement, law enforcement mustadjust and adapt unless the situation is one in which there is a realistic possibility of aconstitutional amendment,20 or law enforcement must wait until there is either a change inthe membership of the court or a carefully prepared test case—or both—to revisit thequestion.

Initial Contacts and the Fourth Amendment

The Fourth Amendment prohibits “unreasonable” searches and seizures. Thus, if defensecounsel can convince the court that an unreasonable search or seizure has occurred, thecharges are most likely to be dismissed. As a result, there is a considerable body of case lawon questions such as whether the initial contact between an officer and a suspect involved a“search” or “seizure.” If the officer seized the driver, the prosecution must be prepared tojustify the seizure and demonstrate that it was “reasonable.” However, if no seizureoccurred, no justification is required: the Fourth Amendment does not forbid officers fromwalking along a street or through a parking lot looking into every vehicle, or from engagingin ordinary conversations with anyone in the vehicle or standing on the sidewalk.

Approaching Stopped Vehicles

Generally, it is not a seizure to approach a person or vehicle in a public place. Thus, if anofficer approaches people sitting in a parked vehicle and then observes open alcoholcontainers in the vehicle or other contraband, the act of approaching the people is not aseizure, and the seizure that follows the observation of evidence of criminal activity isjustified.21 However, if the officer, instead of getting out of the squad car and approachingthe person, summons the person to the squad car, a seizure has occurred.22

Likewise, if an officer responding to a call inadvertently blocks a vehicle’s futuremovement, then no seizure occurs.23 However, if an officer intentionally parks his squadcar in a position that prevents the vehicle from moving, then a seizure occurs.24

18 See California v. Hodari, 499 U.S. 621, 111 S.Ct. 1547 (1991).19 See Welfare of E.D.J., 502 N.W.2d 779 (Minn. 1993).20 Thus, when the court held, in State v. Hamm, 423 N.W.2d 379 (Minn. 1988), that six-person juries in misdemeanorand gross misdemeanor cases violated the state constitution, although statutes have provided for six-person juries sincethe 1851 Territorial Statutes, the bill to amend the constitution—which easily passed and was overwhelminglyapproved at the next election—was introduced before the decision was issued—suggesting that someone at the courttipped someone at the legislature of the impending decision, which would provide the necessity for the legislation.Incidentally, Mr. Hamm was tried again for DWI with a 12-person jury and was convicted.21 See, e.g., State v. Alesso, 328 N.W.2d 685 (Minn. 1983); State v. Vohnoutka, 292 N.W.2d 756 (Minn. 1980).22 See State v. Day, 461 N.W.2d 404 (Minn. Ct. App. 1990).23 See Erickson v. Commissioner of Public Safety, 415 N.W.2d 698 (Minn. Ct. App. 1987) (officers answering a callabout persons disrupting a wedding party parked their squad cars in the driveway, blocking egress, and entered thebuilding to discover that their suspects were now seated in a car they had inadvertently blocked in the driveway).24 See, e.g., Klotz v. Commissioner of Public Safety, 437 N.W.2d 663 (Minn. Ct. App. 1989); State v. Sanger, 420N.W.2d 241 (Minn. Ct. App. 1988).

Watne I-5

In a number of jurisdictions, there are cases that refer to the “communitycaretaker function” of the police officer, under which officers offer help to citizens in aninfinite number of circumstances.25 While Minnesota decisions have not used the term, ithas long been recognized as a legitimate basis of police contacts with individuals. Thus,police may stop to check on a vehicle stopped along the highway to see if the occupantsneed assistance.26

Because police have a duty to investigate motor vehicle crashes, no furtherjustification is needed to approach crash scenes and get information from drivers andwitnesses.

Stopping Moving Vehicles

It is a seizure for an officer to stop a moving vehicle, and that seizure must be justified.One formulation is that the officer must have a particularized and objective basis forsuspecting the particular person who has been stopped of criminal activity.27 TheMinnesota Supreme Court has long adhered to another formulation, articulated in Peoplev. Ingle, 36 N.Y.2d 413, 420, 369 N.Y.S.2d 67, 74, 330 N.E.2d 39, 44 (1975):28

It should be emphasized that the factual basis required to support a stop for a “rou-tine traffic check” is minimal. An actual violation of the Vehicle and Traffic Lawneed not be detectable. For example, an automobile in a general state of dilapida-tion might properly arouse suspicion of equipment violations. All that is required isthat the stop be not the product of mere whim, caprice, or idle curiosity. It is enoughif the stop is based upon “specific and articulable facts which, taken together withrational inferences from those facts, reasonably warrant [the] intrusion.” Terry v.Ohio, 392 U.S. 1, 21, 88 S.Ct. 1868, 1880, 20 L.Ed.2d 889, 906 (1968).

Obviously, if the officer observes an actual traffic violation, the stop is almostalways justified.29 Thus, it is now well established that even if an officer uses a minor

25 See, e.g., State v. Smith, 4 Neb. App. 219, 540 N.W.2d 374 (1995) (under appropriate circumstances a lawenforcement officer may be fully justified in stopping or contacting a vehicle to provide assistance as part of theofficer’s “community caretaker function” without needing any reasonable basis to suspect criminal activity, citing Cadyv. Dombrowski, 413 U.S. 433, 93 S.Ct. 2523 (1973)); State v. Halfmann, 518 N.W.2d 729 (N.D. 1994) (officer wasacting in a community caretaker capacity when he stopped behind the defendant’s stopped car, turned on amberwarning lights, approached the defendant and questioned her; there was no “stop” for Fourth Amendment purposes);State v. Ellenbecker, 464 N.W.2d 427 (Wis. App. 1990) (police officer’s stop to assist a motorist stopped on the roadwith hood raised is not an “investigatory stop” requiring suspicion of illegal activity).26 See, e.g., State v. Compton, 293 N.W.2d 372 (Minn. 1980). See also Paulson v. Commissioner of Public Safety, 384N.W.2d 244 (Minn. Ct. App. 1986); Kozak v. Commissioner of Public Safety, 359 N.W.2d 625 (Minn. Ct. App. 1984);Blank v. Commissioner of Public Safety, 358 N.W.2d 441 (Minn. Ct. App. 1984).27 See, e.g., State v. Kvam, 336 N.W.2d 525 (Minn. 1983), citing United States v. Cortez, 449 U.S. 411, 101 S.Ct. 690(1981) and Delaware v. Prouse, 440 U.S. 648, 99 S.Ct. 1391 (1979).28 See, e.g., State v. Johnson, 444 N.W.2d 824 (Minn. 1989); Marben v. Commissioner of Public Safety, 294 N.W.2d697 (Minn. 1980); State v. Barber, 308 Minn. 204, 241 N.W.2d 476 (1976); State v. McKinley, 305 Minn. 297, 232N.W.2d 906 (1975).29 See, e.g., State v. Bissonette, 445 N.W.2d 843 (Minn. Ct. App. 1989) (failure to signal turn even if no other vehicle isnear enough to create a danger); Holm v. Commissioner of Public Safety, 416 N.W.2d 473 (Minn. Ct. App. 1987)(failure to dim headlights); Daly v. Commissioner of Public Safety, 405 N.W.2d 489 (Minn. Ct. App. 1987) (illegalhonking); State v. Pierce, 347 N.W.2d 829 (Minn. Ct. App. 1984) (loud muffler). However, Minnesota is one state thathas made seat belt law violations a “secondary violation” rather than a “primary violation,” and a stop may not be madesolely on the basis of a seat belt violation. See State v. Fiebke, 554 N.W.2d 755 (Minn. Ct. App. 1996).


offense as a “pretext” for seeking evidence of a more serious offense, the Constitutionis not offended.30 However, in those states in which Indian reservations are located,depending upon the status of a particular reservation, the state may or may not be ableto enforce traffic laws against members of Indian tribes on Indian lands.31

A stop can be valid even if the officer does not actually see the violation, butobserves facts from which the officer can reasonably infer that the violation occurred.32

Likewise, an officer may stop a vehicle seen near the scene of a reported crime to “freezethe situation” even without information linking the crime to any vehicle.33 An officer mayalso validly stop a vehicle seen at unusual times in an area with a history of criminalactivity or a potential for criminal activity.34

An officer who reasonably suspects that the driver of a vehicle does not have avalid driver’s license can validly stop the vehicle,35 even if the suspicion later proves tobe mistaken.36 Likewise, suspicion that the vehicle’s registration has expired justifies astop.37 However, in Minnesota, where certain egregious DWI offenders have theirlicense plates impounded but special series registration plates can be issued to allowfamily members with valid licenses to drive, mere operation of a vehicle with specialplates does entitle officers to stop the vehicle to make certain that the person driving is

30 See, e.g., Whren v. United States, 116 S.Ct. 1769 (1996); State v. Everett, 472 N.W.2d 864 (Minn. 1991); State v.Battleson, 567 N.W.2d 69 (Minn. Ct. App. 1997).31 See, e.g., State v. Stone, 572 N.W.2d 725 (Minn. 1997) (Minnesota’s laws regarding speeding (Minn. Stat. § 169.14 (1996)); driver licensing (Minn. Stat. §§ 171.02, 171.27 (1996)); vehicle registration (Minn. Stat. § 168.09 (1996)); seat belt use (Minn. Stat. § 169.686 (1996)); child restraint seats (Minn. Stat. § 169.685, subd. 5(1996)); motor vehicle insurance (Minn. Stat. § 169.797 (1996)); and proof of insurance (Minn. Stat. § 169.791(1996)) are civil/regulatory for purposes of Public Law 280 and may not be enforced); State v. Robinson, 572N.W.2d 720 (Minn. 1997) (underaged alcohol consumption law may be enforced, but failure to yield to emergencyvehicle may not be enforced). While the perceived distinction is between “civil/regulatory” laws and “criminal”laws, it is rather difficult to discern a test by which one determines which traffic laws are civil/regulatory andwhich are criminal, because nearly all traffic laws exist to regulate human behavior by imposing criminal penaltiesfor misconduct.32 See Berge v. Commissioner of Public Safety, 374 N.W.2d 730 (Minn. 1985) (from speed of vehicle, officer inferredthat driver could not have stopped at stop sign screened from view by vegetation).33 See, e.g., Applegate v. Commissioner of Public Safety, 402 N.W.2d 106 (Minn. 1987) (vehicle leaving parkinglot of apartment complex where burglary was reported to be in progress); Purnell v. Commissioner of PublicSafety, 410 N.W.2d 439 (Minn. Ct. App. 1987) (vehicle leaving area of reported assault stopped on suspicion thatperpetrators, victim or witnesses might be inside); State v. Giebenhain, 374 N.W.2d 573 (Minn. Ct. App. 1985)(stop of vehicle seen near a building where volunteer fireman heard a noise and mistakenly reported a possibleburglary).34 See, e.g., State v. Fish, 280 Minn. 163, 159 N.W.2d 786 (1968) (stop of vehicle emerging from parking lot of closedrural bar at 2:30 a.m.; burglary reported after stop); State v. Clifford, 273 Minn. 249, 141 N.W.2d 124 (1966) (stop tosee why driver was circling a closed business at 3:30 a.m.); State ex rel. Ogg v. Tahash, 273 Minn. 187, 140 N.W.2d692 (1966) (stop of car emerging from small-town alley at 2:00 a.m. to see who they were and what they were doing).See also Olmscheid v. Commissioner of Public Safety, 412 N.W.2d 41 (Minn. Ct. App. 1987) (stop of vehicle on dead-end road near closed businesses in early morning hours).35 See, e.g., State v. Pike, 551 N.W.2d 919 (Minn. 1996) (officer can reasonably assume registered owner is drivingunless there is information to the contrary); State v. Duesterhoeft, 311 N.W.2d 866 (Minn. 1981) (deputy sheriffreasonably assumed registered owner was driving and, based on month-old information, was still under revocation).Although the opinion in Duesterhoeft does not mention it, the deputy sheriff has advised this writer that the vehicle wasa pickup with a “topper,” making it impossible to see who was driving.36 See City of St. Paul v. Vaughn, 306 Minn. 337, 237 N.W.2d 365 (1975) (vehicle validly stopped in mistaken beliefthat driver was the actual driver’s brother, who was under suspension).37 See State v. Delaney, 406 N.W.2d 584 (Minn. Ct. App. 1987) (vehicle with faded-looking temporary registrationpermit in window).

Watne I-7

driving lawfully.38 Driving conduct that is not inherently “illegal” can also justify astop.39 Evasive driving conduct that suggests that the driver is trying to avoid anycontact with law enforcement can justify a stop.40

Many investigatory stops are initiated by citizen complaints to police officers. Citizeninformants are presumed to be reliable whether anonymous or identified. Thus, when thecitizen describes an actual violation, the officer is justified in making an investigatory stopeven if the described violation is not observed by the officer.41 However, if the informant isanonymous, and the “tip” itself does not describe a violation but merely arouses suspicion ofa “possible drunk driver,” the officer needs more information to justify the stop.42 Someappellate judges view anonymous citizen complaints with great suspicion and appear toapply a presumption that any complaint by an anonymous citizen is unreliable.43

The DWI Detection Guide

Two decades ago, the National Highway Traffic Safety Administration (NHTSA)commissioned a study to identify and evaluate the types of driving conduct that led to DWIarrests. The researchers studied thousands of DWI arrest reports from agencies in variousparts of the country, and came up with more than three hundred types of driving conduct.Because speeding is too common among sober drivers to be useful in discriminating thedrunk from the sober, speeding was eliminated as a possible “cue” to DWI violations.NHTSA ultimately developed a list of twenty types of driving conduct that were most likelyto involve a driver with an alcohol concentration of 0.10 percent or more.44

38 See State v. Greyeagle, 541 N.W.2d 326 (Minn. Ct. App. 1995). The opinion suggested that the result might bedifferent if the legislature provided that operation of a vehicle with special series registration plates implied consent tobe stopped and checked out at any time. In 1997, the legislature enacted a new Minn. Stat. § 168.0422 (Supp. 1997),which provides: “A peace officer who observes the operation of a motor vehicle within this state bearing special seriesregistration plates issued under section 168.041, subdivision 6, or 168.042, subdivision 12, may stop the vehicle forthe purpose of determining whether the driver is operating the vehicle lawfully under a valid driver’s license.” As ofthis writing, the appellate courts have not had occasion to rule on the validity of this provision.39 See, e.g., State v. Ellanson, 198 N.W.2d 136 (Minn. 1972) (weaving in lane); State v. Engholm, 290 N.W.2d 780(Minn. 1980) (vehicle traveling very slowly and weaving in lane); State v. Wicklund, 295 Minn. 402, 205 N.W.2d 509(1973) (slow moving vehicle with youthful-looking passengers suggested possible curfew violation).40 See, e.g., State v. Johnson, 444 N.W.2d 824 (Minn. 1989) (driver made eye contact with trooper, promptly turneddown a dead-end road, and was stopped when he came out to find out why he sought to avoid police contact); State v.Petrick, 527 N.W.2d 87 (Minn. 1995) (driver immediately turned into a driveway and turned off lights before comingto a stop upon spotting officer’s squad car).41 See, e.g., Marben v. State, Dept. of Public Safety, 294 N.W.2d 697 (Minn. 1981) (trooper stopped vehicle on basis ofCB radio complaint from truck driver on I-94 that vehicle behind him had been tailgating him for about 70 miles; notailgating seen); State v. Davis, 393 N.W.2d 179 (Minn. 1986) (officer stopped vehicle on basis of complaint fromperson that driver of specific vehicle had just run a red light); City of Minnetonka v. Shepherd, 420 N.W.2d 887 (Minn.1988) (gas station attendant called to report driver of specific vehicle was intoxicated; no bad driving seen).42 See Olson v. Commissioner of Public Safety, 371 N.W.2d 552 (Minn. 1985) (sheriff’s dispatcher reported 911 call of“possible drunk driver” in specific vehicle; no bad driving seen).43 See, e.g., State v. Hjelmstad, 535 N.W.2d 663, 666 (Minn. Ct. App. 1995) (“the trial court’s analysis misses afundamental point from which our analysis must start, that being the credibility of the informant.”).44 See Harris, Howlett and Ridgeway, The Visual Detection of Driving While Intoxicated, Project Interim Report:Identification of Visual Cues and Development of Detection Methods. Anacapa Sciences, Inc., for Department ofTransportation, National Highway Traffic Safety Administration, January 1979. (NHTSA Technical Reference No. HS 805 051; NTIS No. PB 80 108 327); Harris, Dick, Casey and Jarosz, The Visual Detection of Driving WhileIntoxicated, Final Report: Field Test of Visual Cues and Detection Methods. Anacapa Sciences, Inc., for Departmentof Transportation, National Highway Traffic Safety Administration, April 1980 (NHTSA Technical Reference No. HS805 620; NTIS No. PB 81 133 522).


The final listing of cues was published and distributed to law enforcement agenciesby NHTSA.45 The twenty categories of driving conduct shown to be the most likely cuesto DWI violations are

1. Turning with wide radius;2. Straddling center of lane marker;3. Appearing to be drunk;4. Almost striking object or vehicle;5. Weaving;6. Driving on other than designated roadway;7. Swerving;8. Speed more than 10 mph below speed limit;9. Stopping without cause in traffic lane;

10. Following too closely;11. Drifting;12. Tires on center or lane marker;13. Braking erratically;14. Driving into opposing or crossing traffic;15. Signaling inconsistent with driving actions;16. Slow response to traffic signals;17. Stopping inappropriately (other than in lane);18. Turning abruptly or illegally;19. Accelerating or decelerating rapidly; and20. Headlights off.

Probable Cause for Arrest

Absent a statutory requirement that a peace officer can arrest only for violations committedin the officer’s presence, DWI arrests may be made upon probable cause or somesynonymous term.46 “Probable cause” is essentially a reasonable belief,47 based upon anevaluation of the facts as they appear to an experienced police officer at the time.48 It can bebased on the collective knowledge of all officers of one or more agencies, even if the officermaking the actual arrest does not possess all of the necessary information.49

Courts are also to apply an objective standard to the review of an officer’s probablecause determinations, based upon the totality of the circumstances and the recognition thattrained law-enforcement officers are permitted to make “inferences and deductions thatmight well elude an untrained person,”50 and that “great deference” should be given to the

45 See Guide for Detecting Drunk Drivers at Night, National Highway Traffic Safety Administration, U.S. Departmentof Transportation (DOT HS 805 711, 2nd ed. January 1982).46 See, e.g., State v. Harris, 295 Minn. 38, 202 N.W.2d 878, 881 (1972) (“The phrase ‘reasonable and probablegrounds’ is synonymous with the phrases ‘reasonable cause’ and ‘probable cause.’ ”).47 See Id., 202 N.W.2d at 880.48 See, e.g., State v. Carey, 296 Minn. 214, 207 N.W.2d 529 (1973); State v. Stewig, 281 Minn. 331, 161 N.W.2d 673(1968).49 See, e.g., McLaughlin v. State, 291 Minn. 277, 190 N.W.2d 867 (1971); Rancour v. Commissioner of Public Safety,355 N.W.2d 462 (Minn. Ct. App. 1984); State v. Jensen, 351 N.W.2d 29 (Minn. Ct. App. 1984).50 See State v. Kvam, 336 N.W.2d 525 (Minn. 1983) (quoting from United States v. Cortez, 449 U.S. 411, 418, 101S.Ct. 690, 695 (1981).

Watne I-9

officer’s judgment.51 Since the “totality of the circumstances” is the proper test for probablecause, a court must, of necessity, consider all evidence available to the officer, even if itwould not normally be admissible to prove the offense itself, such as so-called “hearsay”evidence.52 Because the standard for review is objective, it is not necessary that the officerbelieve that probable cause exists53 or express the opinion that probable cause existed.54

Generally, the controversies involve whether the officer has probable cause tobelieve that the person was driving, operating or physically controlling the vehicle, or“under the influence” of alcohol or other drugs, etc. Driving, like any other element, canbe proved by circumstantial evidence55 or admissions,56 as well as by the officer’s directobservation.

Because individuals vary widely in their reactions to alcohol and to other drugs,there are many possible symptoms or clues to impairment. In Minnesota, the courts haverejected the claim that some particular “laundry list” of symptoms is necessary to proveimpairment and have held that even one physical indicator can be sufficient.57 However,there must be at least one physical indicator of alcohol use.58 Not even the odor of analcoholic beverage is an indispensable element, even though usually present.59

51 See State v. Olson, 342 N.W.2d 638 (Minn. Ct. App. 1984).52 Technically, hearsay evidence is evidence of an out-of-court statement that is offered to prove the truth of the matterasserted and does not apply to an out-of-court statement offered to prove that the officer had a reasonable basis forbelieving that the suspect committed the offense charged. See, e.g., Minn. R. Evid. 801(c). However, some courtsprefer to use the term “hearsay” for all out-of-court statements, and to hold that “reliable hearsay” is admissible toprove probable cause. See, e.g., Andersen v. Commissioner of Public Safety, 410 N.W.2d 17 (Minn. Ct. App. 1987);Edwards v. Commissioner of Public Safety, 381 N.W.2d 27 (Minn. Ct. App. 1986); Schlemme v. Commissioner ofPublic Safety, 363 N.W.2d 781 (Minn. Ct. App. 1985).53 See State v. Speak, 339 N.W.2d 741 (Minn. 1983).54 See State v. Driscoll, 427 N.W.2d 263 (Minn. Ct. App. 1988); Sarb v. Commissioner of Public Safety, 362 N.W.2d405 (Minn. Ct. App. 1985).55 See, e.g., State v. Pieschke, 295 N.W.2d 580 (Minn. 1980) (Pieschke identified as the driver by two of his threecompanions but denied driving; companions all changed their stories at trial; Pieschke’s conviction affirmed). See alsoState v. Hunt, 356 N.W.2d 801 (Minn. Ct. App. 1984) (Hunt claimed wife was driving at time of accident); Vertina v.Commissioner of Public Safety, 356 N.W.2d 412 (Minn. Ct. App. 1984) (Vertina found lying next to his crashedmotorcycle with broken leg; suggested someone else may have been driving; while the appeal was pending, Vertinawas charged again—when he crashed another motorcycle and broke the same leg).56 See, e.g., Johnson v. Commissioner of Public Safety, 366 N.W.2d 347 (Minn. Ct. App. 1985); Steinberg v. State,Dept. of Public Safety, 357 N.W.2d 413 (Minn. Ct. App. 1984); Hewitt v. Commissioner of Public Safety, 352 N.W.2d75 (Minn. Ct. App. 1984).57 See, e.g., State v. Hicks, 301 Minn. 350, 222 N.W.2d 345 (1974); Holtz v. Commissioner of Public Safety, 340 N.W.2d 363 (Minn. Ct. App. 1983).58 See Musgjerd v. Commissioner of Public Safety, 384 N.W.2d 571 (Minn. Ct. App. 1986) (trooper dealing withmotorcyclist who crashed and severely injured leg on windy night did not look closely for signs of alcohol use,ultimately inferred from circumstances of unexplained crash in good weather on good road late at night that alcoholmust be involved; asked second trooper to get a blood test; and second trooper was not able to get close enough todriver in operating room to check for odor of alcoholic beverage, etc.).59 See State v. Graham, 176 Minn. 164, 222 N.W. 909 (1929) (DWI conviction affirmed although all seven prosecutionwitnesses admitted that they did not smell the odor of an alcoholic beverage on Graham’s breath). See also Johnson v.State, Dept. of Public Safety, 351 N.W.2d 2 (Minn. 1984) (citizen who observed driver park after striking anothervehicle and then walk unsteadily into a restaurant could reasonably believe the driver to be under the influence even ifnever close enough to see bloodshot eyes, hear speech, or smell breath). This can be important in cases where there isother evidence to suggest alcohol impairment but where cold winds or a bad cold prevent the officer from detecting theodor of an alcoholic beverage—and in cases where the officer has no sense of smell, a rather unusual handicap. Thus,in State v. Jensen, 351 N.W.2d 29 (Minn. Ct. App. 1984), the officer who stopped the driver released him when theofficer’s stuffed-up nose prevented him from smelling the odor of an alcoholic beverage, but a DWI arrest was thenmade by a second officer whose nose was in better working order.


While field sobriety tests are not required, they are a useful aid in detectingimpaired drivers. Field sobriety tests, including the horizontal gaze nystagmus (HGN)test are properly admitted as evidence of impairment.60 There is no requirement that aMiranda warning be given before administering field sobriety tests61 or when asking“general on-the-scene questions” at the scene of an accident or traffic stop.62

In addition, a variety of portable breath test (PBT) instruments are available toperform roadside preliminary screening tests. Some statutes limit the evidentiary use ofsuch tests. The original rationale for those limitations appears to have been the questionablereliability of the instruments available at the time statutes were first enacted to expresslyauthorize and encourage the use of the new technology. Later generations of instrumentsare so greatly improved that there seems to be little reason to continue the limitations onevidentiary use, especially since there are no such limitations on the use of officer’sobservations. Thus, in Minnesota, the statute has been amended to expressly authorize theuse of PBT evidence in implied consent litigation, civil actions arising from the use of amotor vehicle, minor consumption cases, and cases involving enforcement of totalabstinence restrictions.63

In Minnesota, it is not necessary to offer either the calibration records of the PBTor testimony from the person who calibrated the instrument to have the results admitted.64

“Investigatory Detention” vs. “Arrest”

A temporary investigatory detention does not necessarily constitute an arrest, eventhough the person is not “free to go,” where the duration of the detention is “reasonable”under all of the circumstances. Thus, detaining two robbery suspects in the back ofseparate squad cars for 61 minutes, during which time the suspects’ shoes were taken andmatched with footprints at the crime scene, did not constitute an arrest.65 Likewise, whena lone trooper trying to deal with several people at the scene of a traffic fatality detainedthe driver for approximately two hours while investigating the incident, and then decidedto “arrest,” had not illegally arrested the driver.66 The distinction between “investigatory

60 See, e.g., State v. Klawitter, 518 N.W.2d 577 (Minn. 1994). The Iowa Supreme Court stated, in rejecting objectionsto the use of HGN test evidence, that “[a]t the outset we note that the principal obstacle to the admissibility of thehorizontal gaze nystagmus test may be its pretentiously scientific name. Though cumbersome, the test’s title is quitedescriptive.” State v. Murphy, 451 N.W.2d 154, 156 (Iowa 1990).61 See Butler v. Commissioner of Public Safety, 348 N.W.2d 827 (Minn. Ct. App. 1984). In Baribeau v. Commissionerof Public Safety, No. C0-87-1883 (Minn. Ct. App. Apr. 19, 1988) (unpublished), the court refused to follow thecontrary Colorado decision in People v. Carlson, 677 P.2d 310 (Colo. 1984).62 See, e.g., State v. Perkins, 353 N.W.2d 557 (Minn. 1984) (Miranda generally does not apply to temporaryinvestigative detentions); Steinberg v. State, Dept. of Public Safety, 357 N.W.2d 413 (Minn. Ct. App. 1984); State v.Kline, 351 N.W.2d 388 (Minn. Ct. App. 1984). One illustration of how much information an officer can elicit at a crashscene is State v. Taylor, 437 P.2d 853 (Ore. 1968), in which the officer obtained admissions that it was Taylor’s car, hewas driving, where he was going, that he had been drinking, what he drank, how much he drank, where he drank, whenhe started, and that he did not presently know where he was.63 See Minn. Stat. § 169.121, subd. 6 (1996). The statute does not address other common uses of PBT instruments, suchas by officers seeing if a passenger is sober enough to drive the car away after the driver has been arrested, by hospitalsand detox facilities screening incoming patients, or by schools trying to exclude persons using alcohol from attendingschool dances.64 See, e.g., Steele v. Commissioner of Public Safety, 439 N.W.2d 427 (Minn. Ct. App. 1989); Lundquist v.Commissioner of Public Safety, 411 N.W.2d 608 (Minn. Ct. App. 1987).65 See State v. Moffatt, 450 N.W.2d 116 (Minn. 1990).66 See Kirsch v. Commissioner of Public Safety, 440 N.W.2d 147 (Minn. Ct. App. 1989).

Watne I-11

detention” and “arrest” can be important because of questions as to whether the officerhad probable cause at the time the arrest occurred, and when the situation became a“custodial arrest” requiring a Miranda warning before further questions.67

Arrests Outside the Officer’s Jurisdiction

In Minnesota, it is well settled that officers may arrest DWI violators outside theirjurisdictions, whether after a pursuit68 or when the officer is outside his jurisdiction in thecourse of duty.69 This is true even if the officer has to cross a state line, whether pursuinga violator70 or following an injured driver to a hospital in a neighboring state.71 Recently acitizen’s arrest in Burnsville, Minnesota, by an off-duty Eau Claire, Wisconsin, officerwho followed a driver from the Mall of America to her residence and detained her until aBurnsville officer could take over, was held to be proper.72

Warrantless Arrests in Dwellings

Because a person cannot defeat an arrest that was initiated in a public place by retreating intoa dwelling,73 officers may pursue traffic offenders into a residence.74 Likewise, a consensualentry into a home permits an in-home arrest.75 “Exigent circumstances” also permitnonconsensual entry into a dwelling.76 However, absent “hot pursuit,” consent or “exigentcircumstances,” a nonconsensual entry into a dwelling is generally considered a violationof the Fourth Amendment’s prohibition against unreasonable searches and seizures.77

67 Generally, a Miranda warning is required only for a “custodial interrogation,” and the brief roadside questioning of amotorist detained for a routine traffic stop is not a custodial interrogation even if the person is seated in the squad carand is not free to leave immediately. See, e.g., Berkemer v. McCarty, 468 U.S. 420, 104 S.Ct. 3138 (1984); State v.Herem, 384 N.W.2d 880 (Minn. 1986); City of Burnsville v. Marsylla, 349 N.W.2d 829 (Minn. 1984).68 See, e.g., Windschitl v. Commissioner of Public Safety, 355 N.W.2d 146 (Minn. 1983) (Springfield officer checkingcity cemetery outside city limits pursued driver seen exiting cemetery after hours).69 See, e.g., State v. Bunde, 556 N.W.2d 917 (Minn. Ct. App. 1996) (city officer several miles outside city observingtraffic violation); Shull v. Commissioner of Public Safety, 398 N.W.2d 11 (Minn. Ct. App. 1986) (Wabasha Countydeputy observed violation in Olmsted County); Bounds v. Commissioner of Public Safety, 361 N.W.2d 145 (Minn. Ct.App. 1985) (Wadena County deputy in Todd County).70 See, e.g., Piotrowski v. Commissioner of Public Safety, 453 N.W.2d 689 (Minn. 1990) (Moorhead, Minnesota officerpursuing semaphore violator into Fargo, North Dakota); Moelter v. Commissioner of Public Safety, 456 N.W.2d 270(Minn. 1990) (Stillwater, Minnesota officer pursuing violator into Wisconsin); State v. Ault, 453 N.W.2d 699 (Minn.1990) (North Dakota officer pursuing violator into Minnesota); Swapinski v. Commissioner of Public Safety, 368N.W.2d 322 (Minn. Ct. App. 1985) (Wisconsin officer pursuing violator into Minnesota); State v. Sellers, 350 N.W.2d460 (Minn. Ct. App. 1984) (same).71 See, e.g., State v. Torgerson, 453 N.W.2d 698 (Minn. 1990) (Moorhead, Minnesota, officer arrested and obtained test athospital in Fargo, North Dakota); Boland v. Commissioner of Public Safety, 520 Minn. Ct. App. 1994) (Minnesota officerinvestigating Minnesota accident could invoke the Minnesota statute and procedures to obtain test at hospital in Fargo).72 See Lamoureux v. Commissioner of Public Safety, No. C6-97-1999 (Minn. Ct. App. Apr. 28, 1998) (unpublished).73 See, e.g., United States v. Santana, 427 U.S. 38, 96 S.Ct. 2406 (1976). See also Costillo v. Commissioner of PublicSafety, 416 N.W.2d 730 (Minn. 1987).74 See, e.g., State v. Paul, 548 N.W.2d 260 (Minn. 1996); State v. Koziol, 338 N.W.2d 47 (Minn. 1983). See alsoSteinbrenner v. Commissioner of Public Safety, 413 N.W.2d 557 (Minn. Ct. App. 1987); Pahlen v. Commissioner ofPublic Safety, 382 N.W.2d (Minn. Ct. App. 1986).75 See, e.g., Carlin v. Commissioner of Public Safety, 413 N.W.2d 249 (Minn. Ct. App. 1987); Overline v.Commissioner of Public Safety, 406 N.W.2d 23 (Minn. Ct. App. 1987); Pesterfield v. Commissioner of Public Safety,399 N.W.2d 605 (Minn. Ct. App. 1987).76 See, e.g., State v. Storvick, 428 N.W.2d 55 (Minn. 1988) (vehicular homicide case).77 See, e.g., State v. Olson, 436 N.W.2d 92 (Minn. 1989). See also Pullen v. Commissioner of Public Safety, 412N.W.2d 780 (Minn. Ct. App. 1987); Krause v. Commissioner of Public Safety, 358 N.W.2d 481 (Minn. Ct. App. 1984).


Post-Arrest Testing

The administration of blood, breath, urine, or other “chemical tests” for alcohol andother drugs is a search and seizure. In Rochin v. California, 342 U.S. 165, 72 S.Ct. 205(1952), the Court concluded that pumping the stomach of a drug offense suspect whohad swallowed the evidence was a search, which shocked the conscience and therebyviolated the Fourth Amendment. However, in Breithaupt v. Abram, 352 U.S. 432, 77S.Ct. 408 (1957), the Court concluded that the withdrawal of blood samples was such acommon medical procedure that taking blood samples for alcohol testing was not shockingto the judicial conscience and did not offend the Fourth Amendment. However, it was not until Schmerber v. California, 384 U.S. 757, 86 S.Ct. 1826 (1966), that the Courtestablished that taking body samples for alcohol testing was not a Fifth Amendment “self-incrimination” issue, but a Fourth Amendment “search and seizure” issue. Thesample sought is physical evidence, not “testimonial” evidence, and may be taken withouta warrant and without consent where there is probable cause to believe that the personcommitted a DWI violation and that the sample will provide evidence to support or refutethe charge.

Meanwhile, before Schmerber, concern that alcohol testing involved FifthAmendment concerns led to early requirements that test evidence be admissible onlyif “voluntarily” obtained. Since few DWI suspects were clamoring to be tested, effortswere made to find ways to induce cooperation in the testing process. New Yorkdevised the “implied consent” concept, by which drivers, by the very act of operatinga motor vehicle, implied consent to testing if arrested for DWI, with a loss of drivingprivileges as a sanction for noncompliance. Other states soon followed suit and, by1973, with federal encouragement, Illinois became the 50th state to enact an impliedconsent law.

However, if the DWI suspect submitted to testing, there would be no loss ofdriving privileges until and unless the person was convicted of the DWI violation. In1976, Minnesota became the first state to use the implied consent mechanisms to revokedriving privileges of DWI suspects who submitted to tests disclosing an alcoholconcentration of 0.10 percent or more. With federal encouragement, this mechanism hasbeen adopted by more than half the states.

By now, the statutory approaches of the 50 states are so varied in the details thatno attempt is made in this paper to address the various means by which test evidence issecured and produced in court. Requirements can be rigid or flexible, depending onjurisdiction and the most recent decisions of the legislatures and appellate courts. Fromthe perspective of enhancing enforcement activities, it appears that the public interest intraffic safety is best served by flexible and non-rigid requirements.

Minnesota is one state that has adopted the more flexible approach. When“chemical test” evidence was first authorized by statute in 1957, it had to be “voluntarily”obtained. After the implied consent law was enacted in 1961, the test had to be obtainedeither voluntarily or pursuant to the implied consent law. In 1984, those restrictions wereeliminated so that any relevant test evidence would be admissible if obtained pursuant tothe implied consent law or any other method permissible under the Constitution. Evenbefore those amendments, the courts had held that a nonconsensual blood test obtainedfrom a driver involved in a fatal crash under authority of Schmerber was admissible in a

Watne I-13

regular DWI proceeding.78 As presently worded, the statute expressly authorizes theadmission of “evidence of the presence or amount of alcohol, controlled substances, orhazardous substances in the person’s blood, breath, or urine as shown by an analysis ofthose items,”79 authorizes the admission of evidence that the person refused to submit totesting,80 and clearly states the nonrestrictive policy of the legislature:81

The preceding provisions do not limit the introduction of any other competentevidence bearing upon the question of whether the person violated this section,including tests obtained more than two hours after the alleged violation andresults obtained from partial tests on an infrared breath-testing instrument. Aresult from a partial test is the assurement obtained by analyzing one adequatebreath sample, as defined in section 169.123, subdivision 2b, paragraph (b).

The “partial test” refers to a single adequate sample in a breath test administeredusing an Intoxilyzer 5000 running Minnesota’s test procedure, which requires twoadequate samples in a single test sequence. The “any other competent evidence” hasresulted in approval of the admission of the results of a single deficient sample82 and ofthe observations of a doctor treating an injured driver.83 The express authorization ofresults of tests administered more than two hours after driving has resulted in the use oftests taken as many as 111⁄2 hours after driving.84

Disincentives to Arrest and Charge

Apart from the relatively low risk of apprehension resulting from the paucity ofenforcement resources available to deal with the total number of impaired drivers on theroads each day and night, in some areas there are additional considerations that operate asa disincentive to vigorous law enforcement.85

78 See State v. Aguirre, 295 N.W.2d 79 (Minn. 1980).79 See Minn. Stat. § 169.123, subd. 2(a) (1996).80 See Minn. Stat. § 169.121, subd. 2(c) (1996). Before South Dakota v. Neville, 459 U.S. 553, 103 S.Ct. 916 (1983),held that admission of evidence in a DWI case that the driver refused testing did not violate Fifth Amendmentprivileges, Minnesota case law forbade any such evidence. See State v. Andrews, 297 Minn. 260, 212 N.W.2d 863(1973). In 1982, the legislature amended the statute to permit evidence of the absence of any test, without any commentor explanation for the absence of a test, with an instruction that no inferences were to be drawn from the absence of testevidence. This was upheld in State v. Willis, 332 N.W.2d 180 (Minn. 1983). Neville was decided before Willis, and ledsome judges to urge that Andrews be immediately overruled. The majority declined to decide an issue that was nottechnically before the court, and the matter has never been raised since: within 24 hours of the Neville decision, severalbills were introduced in the legislature to expressly authorize admission of evidence of test refusals, and the law was soamended. While the U.S. Supreme Court has not had to revisit the issue, the Court of Appeals has stated that Andrewsis no longer controlling law and has no precedential value. See Connor v. Commissioner of Public Safety, 386 N.W.2d242, 245 (Minn. Ct. App. 1986); Abe v. Commissioner of Public Safety, 374 N.W.2d 788, 790 (Minn. Ct. App. 1985).Since then, the Court of Appeals has also rejected the claim that evidence of test refusals violates a state constitutionalprivilege against self-incrimination. See State v. Berge, 464 N.W.2d 595 (Minn. Ct. App. 1991).81 See Minn. Stat. § 169.121, subd. 2(f) (1996).82 See State v. Kieley, 413 N.W.2d 886 (Minn. Ct. App. 1987).83 See State v. Kunz, 457 N.W.2d 265 (Minn. Ct. App. 1990).84 See State v. Jensen, 482 N.W.2d 238 (Minn. Ct. App. 1992).85 The observations in this section are drawn from 25 years of working with police officers and prosecutors throughoutMinnesota, and from discussions with colleagues in other jurisdictions whose experience suggest that these problemsare by no means unique to Minnesota.


If local public opinion does not support vigorous law enforcement, then onecannot expect officers to remain highly motivated to look for impaired drivers. InMinnesota, there have been counties where jurors in DWI cases have routinely acquittedsuspects because they felt that police have no business interfering as long as nobody wasseriously injured while, in an adjacent county, jurors were so hostile to DWI violatorsthat nobody at the courthouse could recall a jury ever coming in with a not guiltyverdict—until one acquittal in recent years. In a jurisdiction where local opinion opposesvigorous enforcement, one cannot expect officers to persist in vigorous enforcement.

If local leadership is uncommitted to enforcement, then the problem for the officeris compounded. In the low-enforcement county referred to in the previous paragraph, themayor of one community was quoted in the media as stating that an alcohol-related triplefatality crash following a wedding was just an unfortunate accident, and that the people inthe community did not believe the driver should be charged with vehicular homicide.86 Inseveral small towns, the city fathers and mothers have fired police officers who actuallysought to enforce DWI and other traffic laws, expressing the position that such heavy-handed enforcement (i.e., stopping people exceeding the speed limit by 15 miles perhour) might be fine in the sinful Twin Cities, but it was unacceptable in their virtuoussmall towns and might drive business from local merchants to competitors in othercommunities. In one community, an officer was fired after he committed the offense ofarresting one of the mayor’s close friends for DWI.87 One sheriff did not want any DWIsuspects booked into his jail and would promptly order his jailers to release them as soonas they called him.88 After a repeat DWI offender who happened to be a personal friendof another sheriff had his license administratively revoked, the sheriff sent a memoforbidding his deputies to report any cases to the Department of Public Safety until theywere reviewed by himself or his chief deputy (his son and successor as sheriff) to see ifsome other disposition would not be more appropriate. In another county, the sheriff wasnot interested in DWI enforcement and made it clear that the State Patrol was notwelcome in his county except to write accident reports on I-90 and U.S. 75. For years, hiscounty was one of three that produced fewer than 10 DWI convictions per year—and theother two had populations below 10,000. In one Twin Cities suburb, an officer whoignored the unwritten law that one does not arrest hometown residents saw the policechief retaliate by first forbidding him to take suspects to a neighboring city for breathtests and then having him repeatedly rewrite his “unsatisfactory” reports. Seeing thehandwriting on the wall, the officer applied for a new job in Minneapolis, where he hasbeen working ever since.89

86 One police officer working in that community later advised this writer that the mayor did not speak for as manypeople as he thought. Ultimately, the driver pleaded guilty to three counts of vehicular homicide on the day his jurytrial was to begin.87 The officer sued for wrongful discharge, won an award in excess of $200,000, and went to work as an officer foranother community.88 One officer complained that when he stopped a speeder on the way back to his city, he looked up to see the DWIsuspect he had just booked into the jail riding by and laughing.89 That officer was the arresting officer in State, Dept. of Public Safety v. Wiehle, 287 N.W.2d 416 (Minn. 1979), inwhich a driver arrested after a hit-and-run crash went into convulsions due to a drug overdose before a formal testrequest could be made. He rushed him to the hospital, returned to the police station for a blood kit, and had the doctordraw the sample. The implied consent statute did not, at that time, authorize taking tests from unconscious drivers. Hisresourceful solution to the problem was approved by the U.S. Supreme Court, and the statute was subsequentlyamended to expressly authorize that procedure.

Watne I-15

The influence of leadership also appears when one compares the approach andresults in Kandiyohi County, about a hundred miles west of the Twin Cities, with those ofits neighboring county to the south. The populations are quite similar in size and makeup,but the Kandiyohi County sheriff made DWI enforcement a priority, while his colleaguedid not. The result was that the Kandiyohi deputies looked for DWI violators, and foundthem easily, while their colleagues to the south seemed to be unable to find DWI violatorsunless they crashed into the squad cars.

Overall, continued public education efforts seem to have paid off in severalrespects. Fewer communities seem prepared to tolerate DWI offenders, and a number ofthose in leadership positions referred to above have been replaced.

In Minnesota, every legislative session since 1971 has revisited the DWI problem,sometimes with several bills. In the past decade, the various proposals have tended to becombined into an omnibus DWI bill attempting to address the DWI laws applicable tomotor vehicles, watercraft, ATVs, aircraft, etc. As of the 1997 amendments, the totalpackage has become so complicated that it is very difficult for anyone to really “get ahandle” on what charges and sanctions apply to a particular offender. Several defenseattorneys, prosecutors, and judges have devised and circulated their own charts in aneffort to provide a quick reference. Despite having looked at the same statutory language,their outlines differ from each other. At a recent meeting between legislators, prosecutors,and law enforcement officers, an officer presented a copy of the driving record of aperson he had recently arrested to a legislator and challenged the legislator to determinewhether the offense was a misdemeanor, gross misdemeanor, or enhanced grossmisdemeanor, whether plates would be impounded or the vehicle forfeited, whether thesuspect could be released to a sober party or must be jailed until maximum bail had beenposted, etc.

When the statute becomes so complex that judges and prosecutors must scratchtheir heads in puzzlement in attempting to apply the statute, the problem is even moreacute for the officer, who must decide now and cannot put off the decision for hours,days, or weeks while an assistant or law clerk researches the question. Thus, at theaforementioned meeting, the officers strongly supported making the statute more simple,so that all concerned can understand and apply it more easily.

Another major complaint of officers is the complexity and amount of paperworkinvolved in even routine DWI arrests, let alone those involving recidivists. At the samemeeting mentioned above, the officer emphatically illustrated his point by presenting thelegislator with a series of about 15 documents the officer had to fill out dealing with everyfacet of the arrest, field sobriety tests, implied consent testing, vehicle inventories, plateimpoundment, authority to detain, etc. Some of the same data—name, address, date ofbirth, driver’s license number, date of incident, time, etc.—had to be duplicated on formafter form. The forms do not have a common design, and must be filled out by hand, oneby one, sometimes taking several hours.

Because each department may design its own forms, there can be significantdifferences that make it more difficult and time-consuming for a judge or a lawyer to dealwith cases initiated by dozens of municipalities, the sheriff’s office, and the state patrol.Officers strongly favor finding ways to simply the paperwork by reducing the number offorms, making the design more consistent and easy to follow, and automating the forms.Some officers have access to computerized forms which, if they can type reasonably well,


both speed up the process and provide a more legible product. This is one area wherestatewide coordination between agencies could result ultimately in greatly reducing thetime needed to complete the paperwork in individual cases.

Another common complaint of officers is the number of hearings that officersmust be prepared to attend—only to have them rescheduled at the last minute if anyoneother than the officer has a conflict or to have the charge dismissed if the officer fails toappear for any reason, even if it is because the officer is on the witness stand in thecourtroom next door. When an officer shows up for the sixth time, ready to testify onwhat would otherwise be a day off, and the matter is postponed again, even enthusiasticofficers may become discouraged. If the officer also works nights, the repeated disruptionof sleep and work schedules imposes a toll that leads some officers to either quit lookingfor impaired drivers or simply refuse to appear repeatedly for 9:00 a.m. hearings afterworking all night. Using the given capabilities of computers, it would seem possible forcourt administrators to coordinate court schedules with work schedules so that officersdoing traffic enforcement at night can get a morning’s sleep and appear in court in theafternoon, while officers working days can appear for court in the morning.

HOW DO VIOLATORS AVOID CONSEQUENCES?

Witnesses Failing to Appear

Officers commonly complain that, if the suspect repeatedly fails to appear, the matter isrescheduled and a bench warrant issued for the suspect’s arrest. However, whenprosecution witnesses fail to appear just once, the charge is simply dismissed. InMinnesota, when implied consent law driver’s license revocation hearings are held in thecourts, the common reason that the violator avoids the consequences mandated by thestatute is that one or more essential witnesses fail to appear.

In some cases, the witness has not received notice of the hearing, a problem thatcan be addressed by examining notification systems to ensure that notices actually get tothe witnesses.90 In each case where a witness fails to appear, a check should be made todetermine why that happened, and efforts should be made to find ways to prevent thatproblem in the future. Thus, when some medical personnel refused to appear in responseto letters requesting their presence, formal subpoenas enforceable by the contempt powerof the court secured their attendance at subsequent hearings.91

Insufficient Police Training

While full-time traffic officers tend to be relatively well trained, when laws become toocomplex, even the most diligent and best-trained officers may not be able to remember tolook for and document all elements necessary for a successful case. General patrol

90 This writer has had a number of cases in which a police department assured the prosecution that the officer hadreceived the notice only to find out, too late, that somebody merely put the notice in the officer’s box at the station—while the officer was on vacation.91 In one Ramsey County case some years ago, a physician refused to appear in court. The judge sent the sheriff’sdeputies to arrest him, and he was brought back to the courtroom, protesting loudly, in handcuffs. The judge informedhim: “Doctor, you have to understand that medicine is not the only game in town.”

Watne I-17

officers, who do not have as much training in DWI enforcement, have a greater problemknowing what to look for, how to find it, and how to document it. In some cases, theofficer may lack even rudimentary training in what to look for and may be unable to findevidence to establish probable cause for arrest.

The obvious remedy is additional training. Most officers are interested inimproving their knowledge and skills and welcome the opportunity to get additionaltraining so that they can avoid making mistakes that bring the prosecution to anunsuccessful conclusion.

Inability of Officers to Adequately Articulate What Is Known

Some cases are lost because an officer is unable to articulate on the witness standinformation actually in his or her possession. Thus, even though time of day, day of theweek, and location form a part of the officer’s consideration in investigating a suspectedDWI violation, many officers forget to articulate that the improper driving conduct wasobserved as the driver left the parking lot of a bar at bar closing time on a weekend night.This is sometimes due to inexperience and lack of training, a problem that can beremedied by experience and additional training.

Gaps in Written Reports

A common defense tactic is to ask the officer if he or she is trained to write complete andaccurate reports, containing “everything important” they know about the case. If theofficer answers in the affirmative the attorney may put on a time-consuming show askingabout details never mentioned in testimony or in the reports. If the officer answers in theaffirmative and testifies to details the officer recalls but did not include in the report, thenthe questions can become more hostile, suggesting that the officer is embellishing orfabricating.

One way to address the issue is to train officers to write more detailed reports andinsist that they do so. However, to the extent that this adds to the burden of paperwork,this insistence may have the net effect of discouraging enforcement efforts.

Another way to address the issue—suggested to this writer by a judge and formerprosecutor—is to train officers to answer in the negative when asked if they are trained toinclude everything important in their reports. When prepared to answer that the officerdoes not attempt to record everything important and cannot anticipate what some otherperson might later consider important, then the report is simply a summary or outline ofthe events to refresh the officer’s memory at trial.92

Of course, the more complete the report is, the more helpful it is to the prosecutor.Additionally, the defense attorney is more likely to conclude that the officer andprosecutor are well prepared and that a resolution short of trial is advisable. As a result,the officers who write the most detailed reports are likely to spend less time sittingaround the courthouse waiting to testify.

92 The judge advised that when he prosecuted a case against an attorney he knew would ask an infinite number ofquestions about facts not involved in the case, he had the officer prepared to disagree that the report was intended tocontain “everything important.” The cross-examination, which had been anticipated to go for two hours, came to anabrupt end.


Lack of Preparation by Witnesses And Prosecutors

Whether one is preparing a wall to receive a new coat of paint, preparing the soil to starta new garden, or preparing for trial, better preparation produces better results. In somepolice departments, officers show up for hearings without any copies of their reports,unable to recall the facts of the case until they have read the reports, and expect theprosecutor to provide reports at the time of hearing. Needless to say, such haphazardpreparation is not calculated to produce a high rate of success. Other officers come withtheir reports, have read and reread the reports, and are thoroughly prepared to describewhat happened.

Likewise, prosecutors who put in little effort into preparation cannot expect to haveas good results as those who prepare thoroughly. Not only should the prosecutor becomevery familiar with the facts recorded by the officer, but the prosecutor should also look forgaps in the evidence and communicate with the officer to seek answers. The number ofunpleasant surprises can be significantly reduced with adequate communication.

Negative Judicial Attitudes

Fortunately, the majority of judges are quite conscientious in the performance of theirduty to apply the law fairly. However, certain individual judges may be indifferent tolegitimate public concerns about traffic safety and, in some cases, hostile to acceptedenforcement procedures.93 Where the judge is open to persuasion, better preparation mayovercome insufficient familiarity with the law or moderate bias. Where the judge is verybiased, contemptuous of the appellate courts and unwilling to follow the law, theprosecution must be prepared to remove that judge or to appeal decisions in cases wherean appeal is possible. Beyond that, it may be possible to have other judges help rein in a“loose cannon,” and call the attention of the public to the judge’s unwillingness to followthe law—especially when the judge faces election.

Statutory “Loopholes”

The more complex a law and the more requirements that are imposed on the officer, themore chances there are for an officer to slip up and make a mistake by which the violatorcan escape the consequences of misconduct. While some of the requirements may beconstitutional imperatives, others are not.

Although in Minnesota the DWI law has applied throughout the state rather thanbeing limited to “streets and highways” since 1937, the language of the implied consentlaw once limited its application to streets and highways, precluding its use when thedriver was found in a private parking lot.94 That limitation was soon removed, and sincethen the statute has applied in all places where the DWI law applies: every square inchof the state.

93 One judge advised this writer that it was his view that it was his function as a judge to find ways to circumvent thestatutes. Another judge, when advised that the prosecutor was obligated to enforce the law the legislature enacted,blurted “why did they pass such a damn foolish law?” Others, more subtle, can simply resolve factual issues against theprosecution secure in the knowledge that credibility questions are essentially immune to any risk of reversal on appeal.94 See State, Dept. of Public Safety v. Halverson, 292 Minn. 468, 194 N.W.2d 573 (1972).

Watne I-19

In Minnesota, for several years there have been separate DWI/Implied Consentlaws for most motor vehicles, snowmobiles and all-terrain vehicles, motorboats, andaircraft. While the legislature has been making them more uniform, it has not yet simplyhad one law cover all such vehicles. This leads to some anomalies and potentialproblems. For example, the statutory definition of “peace officer”—the person who mayinvoke the authority of the implied consent law—varies from statute to statute. Thus aconservation officer employed by the Department of Natural Resources (DNR) would bea “peace officer” if the vehicle is a snowmobile, ATV or motorboat, but not if the vehicleis a pickup truck or car. The DNR officer can still make the DWI arrest but lacks theauthority to make the test request.95 Of the several Indian reservations located inMinnesota, the Mille Lacs Band has made agreements with the state that has led to theirofficers being recognized as having all the authority of a municipal police force workingclosely with both the local sheriff’s departments and the state patrol. Mille Lacs Bandofficers may invoke the Minnesota implied consent law, but officers employed by otherreservations or by the Bureau of Indian Affairs may not. Likewise, officers of variousfederal agencies can arrest for a DWI violation, at least as “citizens,” but they are not“peace officers” for the purposes of the implied consent law.

Likewise, the information required by statute to be provided to the suspect hasdiffered depending on the type of vehicle involved. This can present problems when anofficer arrests a drunken pilot for flying under the influence of alcohol and, because noaircraft Implied Consent Advisory form is available, uses the standard form used fordrivers of cars, trucks, and other land vehicles. Does the reading of the “wrong” advisoryform require suppression of all evidence resulting from the test request?

Other problems can arise when statutory amendments change the informationrequired, and the officer uses an outdated form. Is it fatal that the form did not contain theprecise information required by the amended law?

In Minnesota, there is a special statutory restriction regarding the choice of tests.The officer gets to choose what test to offer, but if that choice is either the blood test orthe urine test, no action may be taken against the person for refusing the blood test unlessan alternative test was also offered. Likewise, no action may be taken against the personfor refusing the urine test unless an alternative test was also offered. If the officer forgetsto offer an alternative test, there is no test—and no driver’s license revocation.Furthermore, the person cannot be charged with the separate crime of refusing to submitto testing under the implied consent law.

At one time, the statutory language required that a blood test be offered in allcases.96 The argument was then made that it was fatal to offer a choice of all threeavailable tests at once, instead of offering an alternative test only after the blood test wasrefused, a claim rejected by the court.97 Likewise, the court rejected the contention that

95 At one time, the definition of peace officer included those employed by a “municipality,” but the license revocationwas rescinded because the officer was employed by a “township” and, thus, was not a peace officer. See State, Dept. ofHighways v. O’Connor, 289 Minn. 243, 183 N.W.2d 574 (1971). At the time, while part-time untrained officers couldwear a badge, carry a gun, and make a DWI arrest, only a full-time officer with special training in traffic law wasauthorized to make the test request. See State, Dept. of Highways v. Halvorson, 288 Minn. 424, 181 N.W.2d 473(1970). Those particular restrictions are long gone, but the statutes have not yet been broadened to include all lawenforcement personnel who may arrest for criminal offenses.96 See State, Dept. of Highways v. McWhite, 286 Minn. 468, 176 N.W.2d 285 (1970).97 See State, Dept. of Highways v. Cornelius, 289 Minn. 521, 184 N.W.2d 779 (1971).


because all three tests were theoretically “available,” the officer must offer all threestatutory tests rather than a blood test and one alternative test.98 Currently, if the officeroffers only a blood test, or only a urine test, there is no problem unless the person refusesto submit to that test.99

In each state, there are probably features that provide “loopholes” or opportunitiesfor mistakes. In some states, there are very rigid requirements on how test evidence isobtained or handled, or on how test evidence is presented in court. This can involve suchmatters as requiring a specific number of seals on a blood test kit or requiring documentaryevidence in every case that a breath test operator is qualified to administer the test insteadof simply having the operator testify to his or her qualifications. Whether there are six orseven seals on a blood test kit does not appear to have any genuine bearing on the questionof whether the sample is what it purports to be or whether the analysis of the sampleproduced a reliable result. Likewise, requiring documentary proof of an operator’squalifications in every case would appear to be a waste of judicial time and resources,offering evidence on matters about which there can be little serious dispute.

RECOMMENDATIONS FOR CHANGE

Employing Existing Technology

In some departments, only a few officers are equipped with radar or laser speedmeasurement instruments. Because illegally high or unusually low speeds are quitefrequently associated with DWI arrests, equipping more officers with newer models ofradar or laser instruments can help increase the number of DWI arrests, as well asgenerally aid enforcement of the speed laws.

Equipping squad cars with video cameras and officers with tape recorders can alsobe helpful. Those devices provide a neutral silent witness that is very persuasive and mayresolve or eliminate potential factual disputes. Some officers do not like them because (a)if the officer does not do a good job, there will be a record of it, and (b) suspects with high tolerance may not look as impaired as HGN and other evidenceindicates. The former objection seems to be an excuse for those who do not aspire todoing the best possible job. The latter objection seems to have some validity. Somepersons with high tolerance may not manifest the physical indicia of alcohol use thatlaypersons might expect to see.100 Likewise, the videotape is not likely to be of that muchassistance in showing how a suspect performed on an HGN test, what the eyes lookedlike, or what the breath smelled like. Unless the equipment is of top quality and used

98 See State v. Boland, 299 Minn. 198, 217 N.W.2d 491 (1974).99 See Workman v. Commissioner of Public Safety, 477 N.W.2d 539 (Minn. Ct. App. 1991).100 In the early 1970s, Minneapolis officers arrested a “ high iron” worker who tested about .30 alcohol concentrationbut, when asked to “walk the line” on camera, did so perfectly—on his hands. About a decade ago, this writerencountered a woman who resolved two DWI cases and an implied consent case before Judge L. W. Yost in McLeodCounty. She and her attorney sat around a conference table with the judge, this writer, and two other prosecutors. Shedid not manifest any physical indicia of alcohol impairment. When she stepped outside to confer with her attorney, aHutchinson police sergeant who was present stated, “I know her, and I don’t think she’s straight right now.” All presentwere of the view that he was completely mistaken, and that she was stone cold sober. However, during the pre-sentenceinvestigation, an Intoxilyzer test produced a reading of 0.22 alcohol concentration, and investigation disclosed that shewas drinking Listerine by the quart for its alcohol content.

Watne I-21

properly, the videotape may not even clearly show the items officers look for in fieldsobriety tests or what was said. However, on balance, the best officers appear to favor theuse of those devices and can effectively use videotape and audiotape to collect andpreserve valuable evidence.101 Using these devices during test requests can also be veryhelpful in documenting that the officer gave all required information, properly vindicatedall of the suspect’s rights, and bent over backward to be reasonable, while the suspectwas either abusive or feigning cooperation with the testing process. In cases where thesuspect goes on the offensive and accuses the officers of sexual misconduct or policebrutality, the tapes can be invaluable aids in defeating such accusations.102

Portable breath test instruments are also very helpful to officers in detectingimpaired drivers, especially those with high tolerance who do not look obviouslyimpaired. It would be helpful if each officer on patrol had a portable breath testinstrument in the squad car so that it would not be necessary to summon some otherofficer to the scene who has one—or release the driver because the officer does not have aportable breath test instrument and cannot readily borrow one. Equipping each officerwould also tend to reduce the likelihood of having to call an extra witness to testify aboutthe portable breath test.

With more officers being equipped with computers, a certain amount of data entrycan be automated, reducing the time that an officer must spend completing reports. Driver’slicenses can have magnetic stripes containing identification data that can be read bycomputers, eliminating the need to manually enter each item repeatedly on numerous forms.

In addition, officers whose squad cars are equipped with mobile data terminalscan run registration and record checks on the fly, enabling an officer to run checks onmany vehicles per hour. These checks can lead to the apprehension of many driverswhose licenses have been revoked—and who may turn out to be intoxicated.

Improved Training in Detection of Impaired Drivers

In every department, continual efforts should be made to train officers in detectiontechniques and to update that training periodically.

As a starting point, each officer should become familiar with sources ofinformation such as the DWI Detection Guide published by NHTSA. Experienced

101 For example, when officers track down a hit-and-run DWI suspect at home, those who grant consent to the officer’sentry commonly insist at the time of trial that there was no consent whatever, confident that it is their word against thatof the officer. Since the state has a relatively high burden of proof as to the legality of the warrantless entry, theiraltered recollections can result in the exclusion of critical evidence and the dismissal of charges. However, this writerhad one case where an Edina officer turned on his tape recorder before knocking on the door, and taped the subsequentconversation with the suspect and his wife. When the suspect, an inebriated lawyer, kept yelling that the officer had “nojurisdiction,” the tape clearly showed his wife telling him “But honey, I invited him in”—twice. This eliminated theanticipated challenge to the warrantless entry.102 For example, in Pastuszak v. Commissioner of Public Safety, No. CX-88-1383 (Minn. Ct. App. Jan. 10, 1989)(unpublished), the driver claimed his refusal was reasonable because he was in excruciating pain from a broken leg. In25 minutes of videotape, he showed no signs of pain, but demonstrated extreme rage, screaming obscene insults at allofficers, threatening to sue the officers for millions, refusing to listen to the test request, kicking at the door when theyleft him alone in the room to cool off, refusing to cooperate with paramedics called to check out his “broken leg,” etc.This writer has been advised of another situation in which a defense attorney who believed his client’s story of policebrutality, preserved on tape, came before a judge to express his great outrage, only to have the prosecutor produce avideotape instead of audiotape. The videotape showed the suspect hurling himself at walls and screaming “don’t hitme!” as the officers watched.


officers should also share tips on detection methods, some of which may not be readilyapparent to a casual observer.103

Officers should also be trained to systematically gather their evidence. It is helpfulto be consistent in one’s procedures in case after case so that defense counsels have lessopportunity to create a “reasonable doubt” as to what the officer did in any particular case.The fact that the officer consistently does certain things in a certain manner or sequenceenhances the officer’s credibility and the probative value of that officer’s testimony.

The use of standardized field sobriety testing can be very helpful. Jurors whoselegal education has come from television, movies, and folklore are aware that fieldsobriety tests exist and expect to hear such evidence. If the officer does not perform them,then jurors may find defense arguing that the officer is not interested in taking the time tothoroughly investigate in the search for truth but merely in running up arrest numbers. Itis helpful if the officer can explain that field sobriety tests are normally administered, butthat they were not administered in this case because (a) the suspect refused to performthem; (b) the suspect was so uncooperative and violent that it was unsafe and pointless toattempt to administer them; (c) the suspect was so obviously intoxicated that it wouldhave been risking injury to attempt to administer them; (d ) weather or road conditions atthe scene made it unsafe to attempt the tests; or because there was some similarjustification for going forward without field sobriety tests. Officers who rely too heavilyon the HGN test or the portable breath test, and who do not take the time to do the fullbattery of tests, should be encouraged to administer the full battery.

Where an alcohol/drug influence report form is used, the officer should take thetime to document as much evidence of impairment as possible, whether it is the standardalcohol-related investigation or a drug recognition evaluation investigation.

Where tests are administered, officers should be trained in the proper methods forcollecting and preserving blood and urine samples and for administering breath tests andshould be discouraged from taking shortcuts. If the officer has followed procedures to theletter, then the credibility of the test evidence is enhanced, and it is less likely that adefense attorney may be able to convince jurors of the existence of reasonable doubt.

Better Training in Avoiding Pitfalls

Where laws are complex, officers need training in the fine points so that they can avoidmistakes in applying the statute that may destroy the case against the violator.

In Minnesota and other states where DWI suspects have a right to consult withcounsel before deciding whether to submit to testing, it is important that officers beadequately trained in what the courts require to vindicate that right.104

103 Trooper Richard Steffen, now a major in the Minnesota State Patrol, demonstrated one of his techniques to thiswriter during a 1974 ridealong. The trooper would take up a position in an adjacent lane on the freeway, slightly behindthe target vehicle, and look at the motion of the front tire. If there were numerous minor steering corrections—even ifthe vehicle was not noticeably weaving—he would stop the vehicle because his experience showed that impaireddrivers commonly make numerous tiny corrections in their efforts to compensate for their impairment. Theinvestigation that followed the stop repeatedly showed his articulated suspicion to be correct.104 In Minnesota, this is complicated by the fact that the appellate courts refuse to provide any “bright line” standards toguide officers, prosecutors, and judges as to how much time the suspect must be allowed, making it impossible for an officerto be certain that what is done will be considered sufficient by the courts. See Kuhn v. Commissioner of Public Safety, 488N.W.2d 838 (Minn. Ct. App. 1992) (the standard is one of “reasonableness” under the specific circumstances of the case).

Watne I-23

Where there is a recognized constitutional or statutory right to additional testing,officers need to be adequately trained as to the nature, extent, and limitations of that right.They need to be trained as to what they must and must not do.

Officers also need training in properly reporting the results of their investigationsand in putting as much detail as possible into the reports. The more detailed the reportsare, the more likely the suspect is going to simply plead guilty instead of attempting tochallenge the charge in court.

Officers also need training in what sort of “tricks” to expect on cross-examination, and how best to deal with them. For example, breath test operators arewell advised not to be gulled into testifying as experts on matters about which theyhave insufficient knowledge. Officers in general should be trained to avoid being ledinto agreeing that their reports contain all important observations, which gives thedefense attorney a chance to question at great length about all the things that mightconceivably happen in a DWI case that do not appear in the reports of this case.

Simplifying Procedures

In every jurisdiction, procedures should be reexamined periodically to find ways tostreamline and simplify them to reduce the time required to process a DWI arrest.How many forms must be completed, how much information each must contain, howthe information is to be entered, are all fertile areas for simplification. How courtshandle DWI cases, how they attempt to coordinate schedules with police work andsleep hours, and how officers are notified of hearing dates are all areas in which itshould be possible to reduce the number of times each person must appear.

Simplifying Statutes

Where annual legislative attention has produced a plethora of complex and sometimesinconsistent provisions that are difficult for officers, prosecutors, and judges tounderstand and apply, the complications tend to be counterproductive. The simpler the statutes, the easier they will be for all to understand and apply, which will increasethe likelihood that officers will make an extra effort to step up enforcement activity,decrease the likelihood of court challenges, and accelerate the handling of those casesthat must be tried.

Setting Priorities with Limited Resources

There are few jurisdictions, if any, with surplus financial resources available to financeevery demand for increased enforcement efforts. Those concerned with the DWI problemmust deal with the reality that the same citizen on the street who favors locking up everyviolator and throwing away the key will balk at being asked to pay taxes to build thenecessary cell blocks. Accordingly, those who seek increased DWI enforcement mustcompete for attention and resources with those whose priorities are in other areas ofenforcement. They must be prepared to make a convincing case based more on factsthan on mere outrage.105


Educating the Public

Ultimately, the greatest reduction in the annual toll of death, injury, and destructionwrought by impaired drivers will probably come from continued efforts to educate thepublic to be less tolerant of impaired drivers. Nationwide efforts by government, MADD,RID, SADD, and other groups in recent decades to educate the public have reducedpublic tolerance for driving while impaired. More people now consider driving whileimpaired to be unthinkable and adjust their behavior accordingly. More people areprepared to intervene to persuade impaired persons not to drive, to take away keys and, ifnecessary, to call the police. Increasingly, drivers equipped with cellular phones areprepared to call 911 to report DWI offenders and to follow the offenders until officers canintercept and stop them. Even bar personnel are prepared to call the police to stop animpaired patron from driving. With these kinds of efforts, one can expect furtherimprovements.

Thus, continued attention by legislatures, government agencies, citizen groupsand the media can increase general deterrence of DWI activity. Public and peer pressuremay also increase specific deterrence of DWI activity of those who are irresponsibleenough to assume a right to drive while impaired, and of those who are unimpressed bythe risk and the consequences of apprehension.

Increasing accurate public information and understanding of the DWI laws canalso have a salutary effect. Potential jurors are drawn from the public. The betterinformed they are, the less likely they are to act on erroneous assumptions, and the morelikely to reach appropriate conclusions.

While nobody has yet found a method to persuade all potential impaired drivers toavoid needless risks of death, injury, and property damage for themselves and others,continued reductions in the annual toll of driving while impaired are attainable. Evenmodest improvements can produce nationwide savings of thousands of lives, tens ofthousands of debilitating injuries, and hundreds of millions of tax and insurance dollars.

105 For example, in preparation for a conference in Colorado, this writer collected figures from Colorado and Minnesota forthe total number of deaths classified as “homicides” and for those attributed to alcohol-related crashes. In both states, yearafter year for the 10-year period examined, drunk drivers killed more people than those using guns, knives, poisons, andother means of homicide. Commonly, drunk drivers killed about twice as many as all “murderers.” Conversations withcolleagues around the country suggest that similar patterns exist in nearly every jurisdiction. Armed with such evidence,one can make a stronger case that since the victims are equally dead and their survivors equally bereft—quite apart frompermanent injuries and property damage—that it is in the public interest to devote at least as many resources to DWIenforcement as to homicide investigation.

Comments on “Prosecuting and Adjudicating DWI Detection Evidence”

KATHRYN STEWART

Pacific Institute for Research and Evaluation

I-25

Mr. Watne has done an excellent job in his paper of indicating how labyrinthine andmind-boggling driving while impaired (DWI) arrest and adjudication systems are andhow vulnerable they are to circumvention by determined offenders and their defenseattorneys. After reading the paper, I could only wonder if anyone ever is successfullyprosecuted. Law enforcement officers, in particular, are required to make Talmudicdecisions about the appropriate grounds for stopping a motorist, when a stop becomes aseizure, who to hold and who to release, whose license plates to impound, etc. All thesedecisions must be made under time pressures and often in adverse circumstances withobstreperous suspects. After an arrest is made, the officer’s work continues, with littlehelp from the system and many frustrating and unreasonable rules. Forms must be filledout and court appearances made, with little regard for efficiency or for other demands onthe officer’s time. After all these efforts, there is still a chance that a weak prosecutor or ajudge with a bad attitude can negate the officer’s best efforts. This is a discouragingpicture, indeed.

In the face of this discouragement, it is essential to keep one thought in mind: Themost important effect of enforcement is its power to deter dangerous and destructivebehavior. Detection, apprehension, and punishment are secondary goals that contribute topublic safety to the degree that they help to maintain a credible threat that deters offense.Of course, in the interests of justice, we want to see offenders punished appropriately.Incapacitation (through the loss of the driving privilege, vehicle, or freedom) helpsprevent further offenses. Also, officers’ morale and enthusiasm for enforcement aredifficult to maintain if their efforts seem constantly to be thwarted. But we as experts andlaw enforcement leaders, as well as line officers, should never lose sight of the fact thatthe most important contributor to deterrence is enforcement. Strategies that amplifydeterrence (e.g., extensive media coverage of enforcement campaigns) have the greatestpotential to improve safety.

Clearly, however, both the public’s safety and its sense of justice would be wellserved by improvements in the system. Mr. Watne has pointed out a number of importantareas that can be streamlined and restructured. He has made some excellent suggestionsthat I will comment on here.

It is obvious that the report-writing, record-keeping, and retrieval systems areridiculously antiquated. This major area of difficulty could be addressed with currentlyavailable hardware and software. A 17-year-old clerk at a 7-Eleven store can swipe mycredit card and know immediately if it is valid and if I am over my credit limit before hesells me a bag of Doritos. There seems no reason why a police officer should not haveimmediate access to the status of a driver’s license when a drinking driver has beenapprehended. Similarly, tax preparation software can be purchased for $29.95 that posesa series of straightforward questions to the taxpayer, performs the necessary calculations,and then prints the correct numbers on whatever tax forms are needed. There seems to be


no reason why similar programs could not be used for arrest reports so that clear, legiblereporting could be done with one simple process. In addition, a report format could beprinted for use in court appearances. This would obviate the kind of problem pointed outby Mr. Watne in which officers sometimes are disorganized or incomplete in theirtestimony.

Mr. Watne points out a second major area of difficulty in the skills, performance,attitudes, and interpretations of the individuals involved in the adjudication process.Breakdowns can occur anywhere in this process, from the inability of an officer to testifyclearly and succinctly up to the realization that an offender happens to be a buddy of themayor. Mr. Watne notes the need for additional training to help alleviate some of thesefailures. This is a very important tool that often is underused. Regardless of how good thetraining is, however, any time a system depends on the performance, judgment, andgoodwill of individual actors, the potential for system breakdown is great.

To the degree that all aspects of the DWI arrest and penalty process are supportedby well-designed systems that operate automatically, fewer breakdowns will occur. Afirst step, of course, is that a major penalty be imposed administratively. Minnesota wasthe first state to adopt administrative license revocation and should be commended forthis pioneering effort. Administrative penalties have much less potential for deliberate oraccidental circumvention. For example in Louisiana, prior to the imposition ofadministrative license revocation, only 56 percent of drivers arrested with blood alcoholconcentrations of 0.10 percent or greater were convicted of DWI. After the law waschanged, almost 90 percent of these arrested drivers received a license penalty.

Clearly, the administrative system should be strengthened and streamlined asmuch as possible. For example, some states minimize the grounds for appeal of a licensepenalty. Some states carry out the appeals process by mail rather than requiring personalappearances by law enforcement officers. If the opportunities for individual failures tocause system failures are minimized, then the the system is likely to run more efficiently.

Finally, and perhaps most important, Mr. Watne makes the excellent point that thesupport of the local community is essential if law enforcement agencies and courts are tobe successful in enforcing driving-while-impaired laws. Many of the most egregiousfailures in the system would be less likely if there were strong community norms againstdriving while impaired, attitudes that support vigorous enforcement and prosecution, andstrong citizen advocacy to bring pressure when necessary. Organizations like MothersAgainst Drunk Driving have used tools such as court watches to excellent effect. Theircontinued vigilance is needed to ensure that the progress that has been made is notundermined. Public outrage can be easily engendered by stories of the multiple offenderwho wipes out a family as they walk to church. The officials who have played a role insuch tragedies can be brought to account. The specter of such an event should hauntevery judge and prosecutor and should guide their actions. It is only with an active,motivated, and aggressive citizenry that we can ensure that individuals and agencies actin the best interests of public safety.

Comments on “Prosecuting and Adjudicating DWI Detection Evidence”

WILLIAM G. SLEVA

Lawrence Superior Court II

I-27

In his paper “Prosecuting and Adjudicating DWI Detection Evidence,” Joel Watneprovides a richly detailed, yet concise, account of the various challenges to virtuallyevery phase of the DWI case, from initial encounter by the officer to trial by jury.

At times he provides a glimpse into the inner workings of the judge, either at thetrial or appellate level. He also provides us with many judges’ decisions after all the facts,arguments, evidence, and personal biases have been weighed, considered, and, at times,disregarded.

The concept of apprehending drivers who are either DWI or per se, testing them,and bringing them to trial seems relatively straightforward and simple. As explained inWatne’s paper, we know that is not the case in some jurisdictions.

Defense lawyers’ creativity in challenging DWI cases is not on the wane.Arguments made today would have seemed ludicrous 10 years ago. What is surprising isthat some of those arguments are successful. I do not say this as a criticism of the defensebar. I expect attorneys to challenge when it is proper to do so. I also expect them torequire the state to jump through every hoop in order to secure a conviction. Our love offreedom and the high regard we place on personal liberty demand it.

With the myriad of challenges based upon constitutional grounds, or challengesmade to specific statutes, adjudication has become so complex that police officers areexpected not only to perform their duties in a professional and competent manner but alsoto be well versed in the intricacies of constitutional law, behavioral and physiologicalscience, psychology, and human relations.

In addition, the police officer must be able to successfully undergo trial by fire inthe courtroom. Never forget that the goal—albeit a short-term one—of the officer andalso of the prosecutor is conviction, whether by plea or by trial.

How do we assist the officer and the prosecutor in reaching this goal and the long-term goal of deterrence? In his paper Joel Watne provides suggestions for improvement. Iwould like to stress two general suggestions: simplification and education.

If one peruses the DWI statutes in the various states, not only will one find lack ofuniformity but also a complexity that defies comprehension at times. I have had to readand reread statutes before I have had any hope of comprehending them—much lessapplying them. I then have had the daunting task of making those statutes intelligible to ajury.

Education must not only be provided to police officers but also to judges,prosecutors, lawyers, and the public. Great strides have been made but much more needsto be done.

I look at the adjudication issue from a narrow perspective. I not only look at it as atrial judge but I also try to look at it as a juror. Our goal should be to get from point A,initial stop, to point B, adjudication, in the most efficient and successful manner possible.Keep in mind that successful does not always mean conviction.


If the law and procedures employed by the officer are straightforward andcomprehensible, then jurors will be better able to decide the case. A frequent complaint Ihear after a trial is that the juror did not understand a certain procedure or that aninstruction was too confusing. Unfortunately, the instructions have to track the statutefairly closely to pass muster with the Court of Appeals.

I would like to add a suggestion not mentioned by Mr. Watne: the creation anduse of standardized alcohol influence reports. These reports would greatly assist anofficer in the field and also in the courtroom. The officer would not have to rely onmemory or practice to ask important questions at the scene but would be prompted by thereport. The answers to the questions would serve to bolster his or her “reasonable stop”and could be used in the courtroom to enhance that officer’s credibility at trial; in myexperience, they have proved invaluable. Many cases come to trial after many months,and an officer’s memory could have had a chance to fade or to blur. With a standardizedreport, the officer could testify that the questions were asked and answered at the scene,and the report would memorialize the conversation the officer had with the accused. Oneof the strategies used by prosecutors at the trial is to convince the jury that the officercould not possibly remember details of a conversation he or she had with the defendantmonths or years ago. If the report is available in the courtroom, with testimony that theanswers were recorded at the scene, this argument is severely weakened.

Keep in mind that some things will always be out of our control, namely,community standards, prosecutors’ goals, and the foibles of judges. With a concertedeffort toward educating the public, police officers, judges, and prosecutors, we willcontinue to make progress in the DWI adjudication arena. Unfortunately, Americansseem to have a short memory. Vigilance and continued emphasis on the dangers ofdrinking and driving are needed in order to further the current trend of reducing deathscaused by drinking and driving.

Documents

Issues and Methods in the Detection of Alcohol and Other Drugs